{"success":true,"data":{"stats":{"total_benchmarks":310},"taxonomy":[{"category":"Disease Biology","suites":[{"id":"dis-targetbench-by-disease-group","name":"TargetBench by Disease Group","benchmarks":[{"id":"dis-targetbench-by-disease-group-target-prediction-for-cancer-diseases","name":"Target Prediction for Cancer Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-cardiovascular-diseases","name":"Target Prediction for Cardiovascular Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-fibrotic-diseases","name":"Target Prediction for Fibrotic Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-inflammatory-immunological-diseases","name":"Target Prediction for Inflammatory & Immunological Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-mental-behavioral-disorders","name":"Target Prediction for Mental & Behavioral Disorders"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-metabolic-endocrine-diseases","name":"Target Prediction for Metabolic & Endocrine Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-neurological-diseases","name":"Target Prediction for Neurological Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-ophthalmic-diseases","name":"Target Prediction for Ophthalmic Diseases"},{"id":"dis-targetbench-by-disease-group-target-prediction-for-reproductive-health-conditions","name":"Target Prediction for Reproductive Health Conditions"}]}]},{"category":"Medicinal Chemistry","suites":[{"id":"med-bench3dfit-molecular-generation-with-3d-conditions","name":"Bench3DFit: Molecular Generation with 3D Conditions","benchmarks":[{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-all-condition","name":"CrossDocked (All-Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-fragment-condition","name":"CrossDocked (Pocket-Fragment Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-only-condition","name":"CrossDocked (Pocket-Only Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-pharmacophore-condition","name":"CrossDocked (Pocket-Pharmacophore Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-all-condition","name":"PLInder (All-Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-fragment-condition","name":"PLInder (Pocket-Fragment Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-only-condition","name":"PLInder (Pocket-Only Condition)"},{"id":"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-pharmacophore-condition","name":"PLInder (Pocket-Pharmacophore Condition)"}]},{"id":"med-fgbench-functional-groups-reasoning","name":"FGBench: Functional Groups Reasoning","benchmarks":[{"id":"med-fgbench-functional-groups-reasoning-functional-group-comparison-direction","name":"Functional Group Comparison Direction"},{"id":"med-fgbench-functional-groups-reasoning-functional-group-comparison-value","name":"Functional Group Comparison Value"},{"id":"med-fgbench-functional-groups-reasoning-functional-group-interaction-direction","name":"Functional Group Interaction Direction"},{"id":"med-fgbench-functional-groups-reasoning-functional-group-interaction-value","name":"Functional Group Interaction Value"},{"id":"med-fgbench-functional-groups-reasoning-single-functional-group-direction","name":"Single Functional Group Direction"},{"id":"med-fgbench-functional-groups-reasoning-single-functional-group-value","name":"Single Functional Group Value"}]},{"id":"med-gpcr-affinity-suite","name":"GPCR Affinity Suite","benchmarks":[{"id":"med-gpcr-affinity-suite-5-ht1a-ic50","name":"5-HT1A IC50"},{"id":"med-gpcr-affinity-suite-5-ht1a-ki","name":"5-HT1A KI"},{"id":"med-gpcr-affinity-suite-5-ht1c-ki","name":"5-HT1C KI"},{"id":"med-gpcr-affinity-suite-5-ht2a-ic50","name":"5-HT2A IC50"},{"id":"med-gpcr-affinity-suite-5-ht2a-ki","name":"5-HT2A KI"},{"id":"med-gpcr-affinity-suite-5-ht2b-ki","name":"5-HT2B KI"},{"id":"med-gpcr-affinity-suite-5-ht5a-ki","name":"5-HT5A KI"},{"id":"med-gpcr-affinity-suite-5-ht6-ic50","name":"5-HT6 IC50"},{"id":"med-gpcr-affinity-suite-5-ht6-ki","name":"5-HT6 KI"},{"id":"med-gpcr-affinity-suite-5-ht7-ki","name":"5-HT7 KI"},{"id":"med-gpcr-affinity-suite-a1-ic50","name":"A1 IC50"},{"id":"med-gpcr-affinity-suite-a1-ki","name":"A1 KI"},{"id":"med-gpcr-affinity-suite-a2a-ic50","name":"A2A IC50"},{"id":"med-gpcr-affinity-suite-a2a-ki","name":"A2A KI"},{"id":"med-gpcr-affinity-suite-a2b-ic50","name":"A2B IC50"},{"id":"med-gpcr-affinity-suite-a2b-ki","name":"A2B KI"},{"id":"med-gpcr-affinity-suite-a3-ki","name":"A3 KI"},{"id":"med-gpcr-affinity-suite-alpha1a-ki","name":"ALPHA1A KI"},{"id":"med-gpcr-affinity-suite-b1-ic50","name":"B1 IC50"},{"id":"med-gpcr-affinity-suite-b1-ki","name":"B1 KI"},{"id":"med-gpcr-affinity-suite-casr-ic50","name":"CASR IC50"},{"id":"med-gpcr-affinity-suite-cb1-ic50","name":"CB1 IC50"},{"id":"med-gpcr-affinity-suite-cb1-ki","name":"CB1 KI"},{"id":"med-gpcr-affinity-suite-cb2-ic50","name":"CB2 IC50"},{"id":"med-gpcr-affinity-suite-cb2-ki","name":"CB2 KI"},{"id":"med-gpcr-affinity-suite-ccr1-ic50","name":"CCR1 IC50"},{"id":"med-gpcr-affinity-suite-ccr2-ic50","name":"CCR2 IC50"},{"id":"med-gpcr-affinity-suite-ccr3-ic50","name":"CCR3 IC50"},{"id":"med-gpcr-affinity-suite-ccr5-ic50","name":"CCR5 IC50"},{"id":"med-gpcr-affinity-suite-crhr1-ic50","name":"CRHR1 IC50"},{"id":"med-gpcr-affinity-suite-crhr1-ki","name":"CRHR1 KI"},{"id":"med-gpcr-affinity-suite-cxcr3-ic50","name":"CXCR3 IC50"},{"id":"med-gpcr-affinity-suite-cxcr4-ic50","name":"CXCR4 IC50"},{"id":"med-gpcr-affinity-suite-d1-ki","name":"D1 KI"},{"id":"med-gpcr-affinity-suite-d2-ic50","name":"D2 IC50"},{"id":"med-gpcr-affinity-suite-d2-ki","name":"D2 KI"},{"id":"med-gpcr-affinity-suite-d3-ki","name":"D3 KI"},{"id":"med-gpcr-affinity-suite-d4-ki","name":"D4 KI"},{"id":"med-gpcr-affinity-suite-ep2-ic50","name":"EP2 IC50"},{"id":"med-gpcr-affinity-suite-ep4-ic50","name":"EP4 IC50"},{"id":"med-gpcr-affinity-suite-eta-ic50","name":"ETA IC50"},{"id":"med-gpcr-affinity-suite-ghs-r-ic50","name":"GHS-R IC50"},{"id":"med-gpcr-affinity-suite-glucagon-ic50","name":"GLUCAGON IC50"},{"id":"med-gpcr-affinity-suite-glucagon-ki","name":"GLUCAGON KI"},{"id":"med-gpcr-affinity-suite-gnrhr-ic50","name":"GNRHR IC50"},{"id":"med-gpcr-affinity-suite-gnrhr-ki","name":"GNRHR KI"},{"id":"med-gpcr-affinity-suite-gpr44-ic50","name":"GPR44 IC50"},{"id":"med-gpcr-affinity-suite-gpr44-ki","name":"GPR44 KI"},{"id":"med-gpcr-affinity-suite-il8rb-ic50","name":"IL8RB IC50"},{"id":"med-gpcr-affinity-suite-machrm1-ic50","name":"MACHRM1 IC50"},{"id":"med-gpcr-affinity-suite-machrm1-ki","name":"MACHRM1 KI"},{"id":"med-gpcr-affinity-suite-machrm2-ki","name":"MACHRM2 KI"},{"id":"med-gpcr-affinity-suite-machrm3-ki","name":"MACHRM3 KI"},{"id":"med-gpcr-affinity-suite-machrm4-ic50","name":"MACHRM4 IC50"},{"id":"med-gpcr-affinity-suite-machrm4-ki","name":"MACHRM4 KI"},{"id":"med-gpcr-affinity-suite-mc3r-ki","name":"MC3R KI"},{"id":"med-gpcr-affinity-suite-mc4r-ic50","name":"MC4R IC50"},{"id":"med-gpcr-affinity-suite-mc4r-ki","name":"MC4R KI"},{"id":"med-gpcr-affinity-suite-mc5r-ki","name":"MC5R KI"},{"id":"med-gpcr-affinity-suite-mglur1-ic50","name":"MGLUR1 IC50"},{"id":"med-gpcr-affinity-suite-mglur2-ic50","name":"MGLUR2 IC50"},{"id":"med-gpcr-affinity-suite-mglur5-ic50","name":"MGLUR5 IC50"},{"id":"med-gpcr-affinity-suite-mglur5-ki","name":"MGLUR5 KI"},{"id":"med-gpcr-affinity-suite-nk1-ki","name":"NK1 KI"},{"id":"med-gpcr-affinity-suite-nk2-ic50","name":"NK2 IC50"},{"id":"med-gpcr-affinity-suite-nk2-ki","name":"NK2 KI"},{"id":"med-gpcr-affinity-suite-nk3-ic50","name":"NK3 IC50"},{"id":"med-gpcr-affinity-suite-nk3-ki","name":"NK3 KI"},{"id":"med-gpcr-affinity-suite-npy1r-ki","name":"NPY1R KI"},{"id":"med-gpcr-affinity-suite-npy2r-ic50","name":"NPY2R IC50"},{"id":"med-gpcr-affinity-suite-npy5r-ic50","name":"NPY5R IC50"},{"id":"med-gpcr-affinity-suite-npy5r-ki","name":"NPY5R KI"},{"id":"med-gpcr-affinity-suite-oprd1-ic50","name":"OPRD1 IC50"},{"id":"med-gpcr-affinity-suite-oprd1-ki","name":"OPRD1 KI"},{"id":"med-gpcr-affinity-suite-oprk1-ic50","name":"OPRK1 IC50"},{"id":"med-gpcr-affinity-suite-oprk1-ki","name":"OPRK1 KI"},{"id":"med-gpcr-affinity-suite-oprm1-ic50","name":"OPRM1 IC50"},{"id":"med-gpcr-affinity-suite-oprm1-ki","name":"OPRM1 KI"},{"id":"med-gpcr-affinity-suite-ox1-ic50","name":"OX1 IC50"},{"id":"med-gpcr-affinity-suite-ox1-ki","name":"OX1 KI"},{"id":"med-gpcr-affinity-suite-ox2-ic50","name":"OX2 IC50"},{"id":"med-gpcr-affinity-suite-ox2-ki","name":"OX2 KI"},{"id":"med-gpcr-affinity-suite-thrombin-ic50","name":"THROMBIN IC50"},{"id":"med-gpcr-affinity-suite-v1a-ic50","name":"V1A IC50"},{"id":"med-gpcr-affinity-suite-v1a-ki","name":"V1A KI"},{"id":"med-gpcr-affinity-suite-v2-ic50","name":"V2 IC50"},{"id":"med-gpcr-affinity-suite-v2-ki","name":"V2 KI"}]},{"id":"med-medchem-holdout","name":"IC50 Prediction","benchmarks":[{"id":"med-medchem-holdout-ic50-prediction-internal-holdout","name":"IC50 Prediction (Internal Holdout)"}]},{"id":"med-ism-benchmarks-admet","name":"ISM Benchmarks: ADMET","benchmarks":[{"id":"med-ism-benchmarks-admet-aggregation-potential","name":"Aggregation potential"},{"id":"med-ism-benchmarks-admet-ames","name":"Ames"},{"id":"med-ism-benchmarks-admet-bcrp-inhibition","name":"BCRP Inhibition"},{"id":"med-ism-benchmarks-admet-blood-brain-barrier-penetration","name":"Blood-Brain Barrier Penetration"},{"id":"med-ism-benchmarks-admet-caco-2-efflux-ratio","name":"Caco-2 Efflux Ratio"},{"id":"med-ism-benchmarks-admet-caco-2-permeability","name":"Caco-2 Permeability"},{"id":"med-ism-benchmarks-admet-clearance-human-hepatocytes","name":"Clearance Human Hepatocytes"},{"id":"med-ism-benchmarks-admet-clearance-human-liver-microsomes","name":"Clearance Human Liver Microsomes"},{"id":"med-ism-benchmarks-admet-clearance-mouse-liver-microsomes","name":"Clearance Mouse Liver Microsomes"},{"id":"med-ism-benchmarks-admet-clearance-rat-liver-microsomes","name":"Clearance Rat Liver Microsomes"},{"id":"med-ism-benchmarks-admet-cyp1a2-inhibition-ic50","name":"CYP1A2 Inhibition IC50"},{"id":"med-ism-benchmarks-admet-cyp2c19-inhibition-ic50","name":"CYP2C19 Inhibition IC50"},{"id":"med-ism-benchmarks-admet-cyp2c9-inhibition-ic50","name":"CYP2C9 Inhibition IC50"},{"id":"med-ism-benchmarks-admet-cyp2d6-inhibition-ic50","name":"CYP2D6 Inhibition IC50"},{"id":"med-ism-benchmarks-admet-cyp3a4-inhibition-ic50","name":"CYP3A4 Inhibition IC50"},{"id":"med-ism-benchmarks-admet-half-life-human","name":"Half Life Human"},{"id":"med-ism-benchmarks-admet-hek293-cc50","name":"HEK293 CC50"},{"id":"med-ism-benchmarks-admet-hepg2-cc50","name":"HepG2 CC50"},{"id":"med-ism-benchmarks-admet-herg-ic50","name":"hERG IC50"},{"id":"med-ism-benchmarks-admet-kinetic-solubility","name":"Kinetic Solubility"},{"id":"med-ism-benchmarks-admet-logd","name":"LogD"},{"id":"med-ism-benchmarks-admet-logp","name":"LogP"},{"id":"med-ism-benchmarks-admet-mdck-mdr1-efflux-ratio","name":"MDCK-MDR1 Efflux Ratio"},{"id":"med-ism-benchmarks-admet-mdck-mdr1-permeability","name":"MDCK-MDR1 Permeability"},{"id":"med-ism-benchmarks-admet-oatp1b1-inhibition","name":"OATP1B1 Inhibition"},{"id":"med-ism-benchmarks-admet-p-gp-inhibition","name":"P-gp Inhibition"},{"id":"med-ism-benchmarks-admet-pampa-permeability","name":"PAMPA Permeability"},{"id":"med-ism-benchmarks-admet-ppb-human","name":"PPB Human"},{"id":"med-ism-benchmarks-admet-ppb-mouse","name":"PPB Mouse"},{"id":"med-ism-benchmarks-admet-ppb-rat","name":"PPB Rat"},{"id":"med-ism-benchmarks-admet-solubility-in-dmso","name":"Solubility in DMSO"},{"id":"med-ism-benchmarks-admet-thermodynamic-solubility","name":"Thermodynamic Solubility"},{"id":"med-ism-benchmarks-admet-vdss","name":"VDss"}]},{"id":"med-ism-benchmarks-protacs-pk","name":"ISM Benchmarks: PROTACs PK","benchmarks":[{"id":"med-ism-benchmarks-protacs-pk-bioavailability-f-upon-oral-administration-in-mice","name":"Bioavailability F (%) upon oral administration in mice"},{"id":"med-ism-benchmarks-protacs-pk-dnauc-dose-normalized-auc-0-t-ngmlhmg-upon-iv-administration-in-mice","name":"DNAUC (dose-normalized AUC, 0-t) (ng/mL·h)/mg upon IV administration in mice"},{"id":"med-ism-benchmarks-protacs-pk-half-life-t12-h-upon-iv-administration-in-mice","name":"Half-life T(1/2) (h) upon IV administration in mice"}]},{"id":"med-kinases-affinity-suite","name":"Kinases Affinity Suite","leaderboard_eligible":false,"benchmarks":[{"id":"med-kinases-affinity-suite-abl1","name":"ABL1"},{"id":"med-kinases-affinity-suite-akt1","name":"AKT1"},{"id":"med-kinases-affinity-suite-akt2","name":"AKT2"},{"id":"med-kinases-affinity-suite-atm","name":"ATM"},{"id":"med-kinases-affinity-suite-atr","name":"ATR"},{"id":"med-kinases-affinity-suite-aurka","name":"AURKA"},{"id":"med-kinases-affinity-suite-aurkb","name":"AURKB"},{"id":"med-kinases-affinity-suite-axl","name":"AXL"},{"id":"med-kinases-affinity-suite-braf","name":"BRAF"},{"id":"med-kinases-affinity-suite-btk","name":"BTK"},{"id":"med-kinases-affinity-suite-bub1","name":"BUB1"},{"id":"med-kinases-affinity-suite-c-met","name":"C-MET"},{"id":"med-kinases-affinity-suite-c-src","name":"C-SRC"},{"id":"med-kinases-affinity-suite-cdk2","name":"CDK2"},{"id":"med-kinases-affinity-suite-cdk7","name":"CDK7"},{"id":"med-kinases-affinity-suite-cdk9","name":"CDK9"},{"id":"med-kinases-affinity-suite-chek1","name":"CHEK1"},{"id":"med-kinases-affinity-suite-csf1r","name":"CSF1R"},{"id":"med-kinases-affinity-suite-csnk1d","name":"CSNK1D"},{"id":"med-kinases-affinity-suite-ddr1","name":"DDR1"},{"id":"med-kinases-affinity-suite-egrf","name":"EGRF"},{"id":"med-kinases-affinity-suite-erbb2","name":"ERBB2"},{"id":"med-kinases-affinity-suite-erbb4","name":"ERBB4"},{"id":"med-kinases-affinity-suite-fgfr2","name":"FGFR2"},{"id":"med-kinases-affinity-suite-fgfr3","name":"FGFR3"},{"id":"med-kinases-affinity-suite-flt3","name":"FLT3"},{"id":"med-kinases-affinity-suite-gsk3b","name":"GSK3B"},{"id":"med-kinases-affinity-suite-igf1r","name":"IGF1R"},{"id":"med-kinases-affinity-suite-ikbkb","name":"IKBKB"},{"id":"med-kinases-affinity-suite-insr","name":"INSR"},{"id":"med-kinases-affinity-suite-irak4","name":"IRAK4"},{"id":"med-kinases-affinity-suite-jak1","name":"JAK1"},{"id":"med-kinases-affinity-suite-jak2","name":"JAK2"},{"id":"med-kinases-affinity-suite-jak3","name":"JAK3"},{"id":"med-kinases-affinity-suite-kit","name":"KIT"},{"id":"med-kinases-affinity-suite-lck","name":"LCK"},{"id":"med-kinases-affinity-suite-map2k1","name":"MAP2K1"},{"id":"med-kinases-affinity-suite-map4k4","name":"MAP4K4"},{"id":"med-kinases-affinity-suite-mapk10","name":"MAPK10"},{"id":"med-kinases-affinity-suite-mapk11","name":"MAPK11"},{"id":"med-kinases-affinity-suite-mapk14","name":"MAPK14"},{"id":"med-kinases-affinity-suite-mapk8","name":"MAPK8"},{"id":"med-kinases-affinity-suite-mapk9","name":"MAPK9"},{"id":"med-kinases-affinity-suite-mapkapk2","name":"MAPKAPK2"},{"id":"med-kinases-affinity-suite-mknk","name":"MKNK"},{"id":"med-kinases-affinity-suite-mknk2","name":"MKNK2"},{"id":"med-kinases-affinity-suite-ntrk-1","name":"NTRK-1"},{"id":"med-kinases-affinity-suite-pdgfr","name":"PDGFR"},{"id":"med-kinases-affinity-suite-pdpk1","name":"PDPK1"},{"id":"med-kinases-affinity-suite-pik3ca","name":"PIK3CA"},{"id":"med-kinases-affinity-suite-pik3cg","name":"PIK3CG"},{"id":"med-kinases-affinity-suite-plk1","name":"PLK1"},{"id":"med-kinases-affinity-suite-plk2","name":"PLK2"},{"id":"med-kinases-affinity-suite-plk4","name":"PLK4"},{"id":"med-kinases-affinity-suite-prkca","name":"PRKCA"},{"id":"med-kinases-affinity-suite-prkdc","name":"PRKDC"},{"id":"med-kinases-affinity-suite-ptk2","name":"PTK2"},{"id":"med-kinases-affinity-suite-ret","name":"RET"},{"id":"med-kinases-affinity-suite-rock2","name":"ROCK2"},{"id":"med-kinases-affinity-suite-rsk2","name":"RSK2"},{"id":"med-kinases-affinity-suite-tek","name":"TEK"},{"id":"med-kinases-affinity-suite-tgfbr1","name":"TGFBR1"},{"id":"med-kinases-affinity-suite-tnik","name":"TNIK"},{"id":"med-kinases-affinity-suite-ttk","name":"TTK"},{"id":"med-kinases-affinity-suite-tyk2","name":"TYK2"},{"id":"med-kinases-affinity-suite-vegfr2","name":"VEGFR2"},{"id":"med-kinases-affinity-suite-zap70","name":"ZAP70"}]},{"id":"med-molecular-mechanics-energy","name":"Molecular Mechanics Energy","benchmarks":[{"id":"med-molecular-mechanics-energy-conformation-ranking","name":"Conformation ranking"},{"id":"med-molecular-mechanics-energy-mmff-energy-prediction","name":"MMFF Energy Prediction"}]},{"id":"med-mumo-multiple-property-optimization","name":"MuMO: Multiple Property Optimization","benchmarks":[{"id":"med-mumo-multiple-property-optimization-bbbp-drd2-plogp","name":"BBBP + DRD2 + pLogP"},{"id":"med-mumo-multiple-property-optimization-bbbp-drd2-plogp-qed","name":"BBBP + DRD2 + pLogP + QED"},{"id":"med-mumo-multiple-property-optimization-bbbp-drd2-qed","name":"BBBP + DRD2 + QED"},{"id":"med-mumo-multiple-property-optimization-bbbp-plogp-qed","name":"BBBP + pLogP + QED"},{"id":"med-mumo-multiple-property-optimization-drd2-plogp-qed","name":"DRD2 + pLogP + QED"}]},{"id":"med-pdbbind-restore-interactions","name":"PDBBind Restore Interactions","benchmarks":[{"id":"med-pdbbind-restore-interactions-3d-interactions-restoration","name":"3D Interactions Restoration"}]},{"id":"med-pkis2-kinase-inhibition-suite","name":"PKIS2 Kinase Inhibition Suite","benchmarks":[{"id":"med-pkis2-kinase-inhibition-suite-pkis2-egfr","name":"PKIS2 EGFR"},{"id":"med-pkis2-kinase-inhibition-suite-pkis2-kit","name":"PKIS2 KIT"},{"id":"med-pkis2-kinase-inhibition-suite-pkis2-ret","name":"PKIS2 RET"}]},{"id":"med-tdc-admet","name":"TDC ADMET","benchmarks":[{"id":"med-tdc-admet-ames-mutagenicity","name":"AMES Mutagenicity"},{"id":"med-tdc-admet-bbb-penetration","name":"BBB Penetration"},{"id":"med-tdc-admet-bioavailability","name":"Bioavailability"},{"id":"med-tdc-admet-caco-2-permeability","name":"Caco-2 Permeability"},{"id":"med-tdc-admet-clearance-hepatocyte","name":"Clearance Hepatocyte"},{"id":"med-tdc-admet-clearance-microsome","name":"Clearance Microsome"},{"id":"med-tdc-admet-cyp2c9-inhibition","name":"CYP2C9 Inhibition"},{"id":"med-tdc-admet-cyp2c9-substrate","name":"CYP2C9 Substrate"},{"id":"med-tdc-admet-cyp2d6-inhibition","name":"CYP2D6 Inhibition"},{"id":"med-tdc-admet-cyp2d6-substrate","name":"CYP2D6 Substrate"},{"id":"med-tdc-admet-cyp3a4-inhibition","name":"CYP3A4 Inhibition"},{"id":"med-tdc-admet-cyp3a4-substrate","name":"CYP3A4 Substrate"},{"id":"med-tdc-admet-dili","name":"DILI"},{"id":"med-tdc-admet-half-life","name":"Half Life"},{"id":"med-tdc-admet-herg-blockers","name":"hERG Blockers"},{"id":"med-tdc-admet-hia","name":"HIA"},{"id":"med-tdc-admet-ld50-acute-toxicity","name":"LD50 Acute Toxicity"},{"id":"med-tdc-admet-lipophilicity","name":"Lipophilicity"},{"id":"med-tdc-admet-p-gp-inhibition","name":"P-gp Inhibition"},{"id":"med-tdc-admet-ppbr","name":"PPBR"},{"id":"med-tdc-admet-solubility","name":"Solubility"},{"id":"med-tdc-admet-vdss","name":"VDss"}]}]},{"category":"Chemical Synthesis","suites":[{"id":"syn-msrs","name":"Multi-step retrosynthesis","benchmarks":[{"id":"syn-msrs-multi-step-retrosynthesis-ursa-drugsclinicals-2026-test-set","name":"Multi-step retrosynthesis, URSA-drugs&clinicals-2026 test set"},{"id":"syn-msrs-multi-step-retrosynthesis-ursa-expert-2026-test-set","name":"Multi-step retrosynthesis, URSA-expert-2026 test set"}]},{"id":"che-reaction-plausibility","name":"Reaction Plausibility","benchmarks":[{"id":"che-reaction-plausibility-reaction-plausibility","name":"Reaction Plausibility"}]},{"id":"che-retrosynthesis-suite","name":"Single-step retrosynthesis, Top-1 task","benchmarks":[{"id":"che-retrosynthesis-suite-single-step-retrosynthesis-ursa-expert-2026-dataset","name":"Single-step retrosynthesis, URSA-expert-2026 dataset"},{"id":"che-retrosynthesis-suite-single-step-retrosynthesis-uspto-50k-test-sample-dataset","name":"Single-step retrosynthesis, USPTO-50k-test sample dataset"}]},{"id":"syn-retrosynthesis-suite","name":"Single-step retrosynthesis, Top-15","benchmarks":[{"id":"syn-retrosynthesis-suite-single-step-retrosynthesis-top-15-task-ursa-expert-2026-dataset","name":"Single-step retrosynthesis, Top-15 task, URSA-expert-2026 dataset"},{"id":"syn-retrosynthesis-suite-single-step-retrosynthesis-top-15-task-uspto-50k-test-sample-dataset","name":"Single-step retrosynthesis, Top-15 task, USPTO-50k-test sample dataset"}]}]},{"category":"Clinical Trials","suites":[{"id":"cli-clinbench-quarterly","name":"ClinBench Quarterly","benchmarks":[{"id":"cli-clinbench-quarterly-all-phases","name":"All Phases"},{"id":"cli-clinbench-quarterly-all-phases-2025q1","name":"All Phases - 2025Q1"},{"id":"cli-clinbench-quarterly-all-phases-2025q2","name":"All Phases - 2025Q2"},{"id":"cli-clinbench-quarterly-all-phases-2025q3","name":"All Phases - 2025Q3"},{"id":"cli-clinbench-quarterly-all-phases-2025q4","name":"All Phases - 2025Q4"},{"id":"cli-clinbench-quarterly-all-phases-2026q1","name":"All Phases - 2026Q1"},{"id":"cli-clinbench-quarterly-phase-1","name":"Phase 1"},{"id":"cli-clinbench-quarterly-phase-1-2025q1","name":"Phase 1 - 2025Q1"},{"id":"cli-clinbench-quarterly-phase-1-2025q2","name":"Phase 1 - 2025Q2"},{"id":"cli-clinbench-quarterly-phase-1-2025q3","name":"Phase 1 - 2025Q3"},{"id":"cli-clinbench-quarterly-phase-1-2025q4","name":"Phase 1 - 2025Q4"},{"id":"cli-clinbench-quarterly-phase-1-2026q1","name":"Phase 1 - 2026Q1"},{"id":"cli-clinbench-quarterly-phase-2","name":"Phase 2"},{"id":"cli-clinbench-quarterly-phase-2-2025q1","name":"Phase 2 - 2025Q1"},{"id":"cli-clinbench-quarterly-phase-2-2025q2","name":"Phase 2 - 2025Q2"},{"id":"cli-clinbench-quarterly-phase-2-2025q3","name":"Phase 2 - 2025Q3"},{"id":"cli-clinbench-quarterly-phase-2-2025q4","name":"Phase 2 - 2025Q4"},{"id":"cli-clinbench-quarterly-phase-2-2026q1","name":"Phase 2 - 2026Q1"},{"id":"cli-clinbench-quarterly-phase-3","name":"Phase 3"},{"id":"cli-clinbench-quarterly-phase-3-2025q1","name":"Phase 3 - 2025Q1"},{"id":"cli-clinbench-quarterly-phase-3-2025q2","name":"Phase 3 - 2025Q2"},{"id":"cli-clinbench-quarterly-phase-3-2025q3","name":"Phase 3 - 2025Q3"},{"id":"cli-clinbench-quarterly-phase-3-2025q4","name":"Phase 3 - 2025Q4"},{"id":"cli-clinbench-quarterly-phase-3-2026q1","name":"Phase 3 - 2026Q1"},{"id":"cli-clinbench-quarterly-phase-4","name":"Phase 4"},{"id":"cli-clinbench-quarterly-phase-4-2025q1","name":"Phase 4 - 2025Q1"},{"id":"cli-clinbench-quarterly-phase-4-2025q2","name":"Phase 4 - 2025Q2"},{"id":"cli-clinbench-quarterly-phase-4-2025q3","name":"Phase 4 - 2025Q3"},{"id":"cli-clinbench-quarterly-phase-4-2025q4","name":"Phase 4 - 2025Q4"},{"id":"cli-clinbench-quarterly-phase-4-2026q1","name":"Phase 4 - 2026Q1"}]}]},{"category":"Longevity","suites":[{"id":"lon-longevitybench-peachy","name":"LongevityBench (Peachy)","benchmarks":[{"id":"lon-longevitybench-peachy-geo-dnam-age-multiclass-lb-0002","name":"GEO DNAm Age / Multiclass (LB-0002)"},{"id":"lon-longevitybench-peachy-geo-dnam-age-pairwise-lb-0006","name":"GEO DNAm Age / Pairwise (LB-0006)"},{"id":"lon-longevitybench-peachy-geo-dnam-age-regression-lb-0010","name":"GEO DNAm Age / Regression (LB-0010)"},{"id":"lon-longevitybench-peachy-gtex-transcriptomic-age-generation-lb-0026","name":"GTEx Transcriptomic Age / Generation (LB-0026)"},{"id":"lon-longevitybench-peachy-gtex-transcriptomic-age-multiclass-lb-0014","name":"GTEx Transcriptomic Age / Multiclass (LB-0014)"},{"id":"lon-longevitybench-peachy-gtex-transcriptomic-age-pairwise-lb-0018","name":"GTEx Transcriptomic Age / Pairwise (LB-0018)"},{"id":"lon-longevitybench-peachy-nhanes-biomarkers-age-multiclass-lb-0030","name":"NHANES Biomarkers Age / Multiclass (LB-0030)"},{"id":"lon-longevitybench-peachy-nhanes-biomarkers-age-pairwise-lb-0034","name":"NHANES Biomarkers Age / Pairwise (LB-0034)"},{"id":"lon-longevitybench-peachy-nhanes-biomarkers-age-regression-lb-0038","name":"NHANES Biomarkers Age / Regression (LB-0038)"},{"id":"lon-longevitybench-peachy-nhanes-biomarkers-mortality-binary-lb-0042","name":"NHANES Biomarkers Mortality / Binary (LB-0042)"},{"id":"lon-longevitybench-peachy-nhanes-biomarkers-mortality-multiclass-lb-0046","name":"NHANES Biomarkers Mortality / Multiclass (LB-0046)"},{"id":"lon-longevitybench-peachy-nhanes-biomarkers-mortality-pairwise-lb-0050","name":"NHANES Biomarkers Mortality / Pairwise (LB-0050)"},{"id":"lon-longevitybench-peachy-olink-proteomic-age-generation-lb-0058","name":"Olink Proteomic Age / Generation (LB-0058)"},{"id":"lon-longevitybench-peachy-olink-proteomic-age-pairwise-lb-0066","name":"Olink Proteomic Age / Pairwise (LB-0066)"},{"id":"lon-longevitybench-peachy-opengenes-expression-binary_masked-lb-0086","name":"OpenGenes Expression / binary_masked (LB-0086)"},{"id":"lon-longevitybench-peachy-synergyage-lifespan-pairwise-ablated-lb-0122","name":"SynergyAge Lifespan / Pairwise (ablated, LB-0122)"},{"id":"lon-longevitybench-peachy-synergyage-lifespan-regression-lb-0126","name":"SynergyAge Lifespan / Regression (LB-0126)"},{"id":"lon-longevitybench-peachy-tcga-cancer-survival-pairwise-lb-0134","name":"TCGA Cancer Survival / Pairwise (LB-0134)"}]}]},{"category":"Therapeutic Biomolecules","suites":[{"id":"str-genbio-hac-hic-suite","name":"GenBio HAC HIC Suite","benchmarks":[{"id":"str-genbio-hac-hic-suite-hac-retention-time","name":"HAC Retention Time"},{"id":"str-genbio-hac-hic-suite-hic-retention-time","name":"HIC Retention Time"}]},{"id":"str-polyreactivity","name":"Polyreactivity","benchmarks":[{"id":"str-polyreactivity-cho-polyreactivity","name":"CHO Polyreactivity"},{"id":"str-polyreactivity-ovalbumin-polyreactivity","name":"Ovalbumin Polyreactivity"}]},{"id":"str-sec-smac","name":"SEC SMAC","benchmarks":[{"id":"str-sec-smac-sec-monomer","name":"SEC %Monomer"},{"id":"str-sec-smac-smac","name":"SMAC"}]},{"id":"str-tm-and-titer","name":"Tm and Titer","benchmarks":[{"id":"str-tm-and-titer-titer","name":"Titer"},{"id":"str-tm-and-titer-tm","name":"Tm"}]}]}],"database":{"dis-targetbench-by-disease-group-target-prediction-for-cancer-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Cancer Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 35 diseases: Acute Leukemia, Acute myeloid leukemia, Biliary tract cancer, Bladder Cancer, Bone sarcoma...","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.4083746348344173,"top_K_average_precision":0.6276723154668927,"moa":1.0773515154972797,"commercial":335.1605658509374,"structure":0.9514837912229883,"druggable":0.7573450490000149,"approval_other":0.3247293152216468,"relevance":109.24294499306025,"bioassay":421.7322006747189,"modulator":13.940280207516867,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","top_K_overlap":0.35535762537149207,"top_K_average_precision":0.5389172892854777,"moa":0.9768323133808636,"commercial":279.90818003009935,"structure":0.8952626333774916,"druggable":0.6532243739290253,"approval_other":0.2663256115064617,"relevance":113.08726777729001,"bioassay":372.7868941228464,"modulator":12.358459151457865,"valid_fraction":1,"rank":4},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.3297538677005661,"top_K_average_precision":0.6078212287249707,"moa":0.9193495321943327,"commercial":320.1343599414052,"structure":0.9076562446833623,"druggable":0.6376776324879968,"approval_other":0.2660719259704463,"relevance":102.27363773029909,"bioassay":353.5313028968165,"modulator":11.339123763167535,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","top_K_overlap":0.33568369572343426,"top_K_average_precision":0.6044625180237106,"moa":0.8894846250477588,"commercial":307.9051695227203,"structure":0.9464134608848395,"druggable":0.6467410299507484,"approval_other":0.28550340501333094,"relevance":115.89574533165705,"bioassay":396.0549629058834,"modulator":12.925799772594374,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.3966934017746635,"top_K_average_precision":0.637788747507184,"moa":1.054993144863978,"commercial":328.58759241350947,"structure":0.9507434417846548,"druggable":0.7548016872769655,"approval_other":0.31241649580549574,"relevance":107.73376072173956,"bioassay":423.36946850807135,"modulator":14.375517436188456,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","top_K_overlap":0.359666460698514,"top_K_average_precision":0.5443541465962248,"moa":0.9472194828824569,"commercial":318.21304649267705,"structure":0.8769980193832889,"druggable":0.680768645352173,"approval_other":0.29657972630701257,"relevance":103.8913447984927,"bioassay":385.71655729398407,"modulator":12.958760978157098,"valid_fraction":0.9428571428571428,"rank":3}],"name":"Target Prediction for Cancer Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-cardiovascular-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Cardiovascular Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 3 diseases: Heart failure, Ischemia, Pulmonary arterial hypertension","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.30071182205328545,"top_K_average_precision":0.5196713757484703,"moa":0.839639751170239,"commercial":94.86779310966811,"structure":0.9264062474866895,"druggable":0.76685325483762,"approval_other":0.4113939423185189,"relevance":53.318708112911416,"bioassay":401.21989255094985,"modulator":11.14982225868198,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","top_K_overlap":0.16368387709851123,"top_K_average_precision":0.2934758539441449,"moa":0.5172810157322353,"commercial":47.61683982683982,"structure":0.6076521739130435,"druggable":0.43327536231884056,"approval_other":0.20092753623188406,"relevance":44.3407043789933,"bioassay":204.50353623188406,"modulator":4.332869565217391,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.25199402562207446,"top_K_average_precision":0.5159564869039345,"moa":0.7715470669235912,"commercial":114.31599657287158,"structure":0.9169582856616825,"druggable":0.7050384003366604,"approval_other":0.25448836811719994,"relevance":61.664073212670246,"bioassay":335.98253836745965,"modulator":5.5551117160479215,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","top_K_overlap":0.2439959261605603,"top_K_average_precision":0.4774889521832742,"moa":0.7061680161809735,"commercial":109.49872384559886,"structure":0.9129835282883744,"druggable":0.7004643908649637,"approval_other":0.3028229489933954,"relevance":65.67968411660951,"bioassay":359.9876556059521,"modulator":7.729634439404791,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.3526180797522261,"top_K_average_precision":0.5521779244598335,"moa":0.9064243128145567,"commercial":174.24376713564212,"structure":0.9055314182355924,"druggable":0.7208717544470721,"approval_other":0.3355942974817748,"relevance":56.467509472455596,"bioassay":365.485746759794,"modulator":8.962406887996725,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","top_K_overlap":0.26654282370745785,"top_K_average_precision":0.4320702903121228,"moa":0.7934558960687009,"commercial":98.72187545093794,"structure":0.8686355831462159,"druggable":0.6601798089200018,"approval_other":0.31812442213710207,"relevance":56.40321330542829,"bioassay":354.7460847972527,"modulator":10.146982609078501,"valid_fraction":1,"rank":3}],"name":"Target Prediction for Cardiovascular Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-metabolic-endocrine-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Metabolic & Endocrine Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 12 diseases: Alcoholic hepatitis, Chronic Kidney Disease, Gestational diabetes mellitus, Hyperlipidemia, Hypertension...","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.3604155912222426,"top_K_average_precision":0.6539932924491473,"moa":1.690293230619169,"commercial":142.5776257570884,"structure":0.8755387619592327,"druggable":0.6622428076171646,"approval_other":0.301179600357581,"relevance":43.8231969504562,"bioassay":277.0113004543992,"modulator":8.374178363838267,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","top_K_overlap":0.27138810088449133,"top_K_average_precision":0.5727311797451023,"moa":1.4045730242155994,"commercial":102.76457618397633,"structure":0.761512143646976,"druggable":0.6010071520235833,"approval_other":0.32161798849672446,"relevance":36.55719122935532,"bioassay":333.4167979952421,"modulator":7.8765995763519845,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.31404808954321967,"top_K_average_precision":0.6628066372950022,"moa":1.5965593794205766,"commercial":142.5776257570884,"structure":0.8748642657739572,"druggable":0.6988308214507869,"approval_other":0.3473393237863221,"relevance":47.657569646087865,"bioassay":255.56780781191387,"modulator":7.524079075142185,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","top_K_overlap":0.30045632593526456,"top_K_average_precision":0.6533065984490877,"moa":1.5478588240349336,"commercial":126.03458693287291,"structure":0.855028541232847,"druggable":0.6727647330203337,"approval_other":0.3477495990193933,"relevance":45.853627601042966,"bioassay":233.17476398843394,"modulator":6.92071188398355,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.37237067323870465,"top_K_average_precision":0.693585537066092,"moa":1.6836844420882453,"commercial":160.7393028800015,"structure":0.8543189146724969,"druggable":0.7244904678126831,"approval_other":0.3732268116932233,"relevance":41.09802327644515,"bioassay":266.72752425311546,"modulator":8.270416748965497,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","top_K_overlap":0.30906290345067894,"top_K_average_precision":0.5126327402872846,"moa":1.5157021576771033,"commercial":106.3186707162588,"structure":0.7412284086623201,"druggable":0.6225086960172219,"approval_other":0.23907146776648502,"relevance":43.732497023408676,"bioassay":207.24511973821802,"modulator":8.646576719982933,"valid_fraction":1,"rank":4}],"name":"Target Prediction for Metabolic & Endocrine Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-fibrotic-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Fibrotic Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 6 diseases: Atherosclerosis, Cystic fibrosis, Idiopathic Pulmonary Fibrosis, Interstitial lung disease, Kidney fibrosis...","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.2575114377439959,"top_K_average_precision":0.4876504457748807,"moa":0.5557212649753346,"commercial":26.074005772005773,"structure":0.8975790875790878,"druggable":0.8650925841834933,"approval_other":0.4592377319650047,"relevance":50.56764814087595,"bioassay":376.4902626666263,"modulator":14.345495716404807,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","top_K_overlap":0.22418754802475735,"top_K_average_precision":0.42057082217487585,"moa":0.6615240619415038,"commercial":53.12189393939394,"structure":0.850227927211455,"druggable":0.768206087742807,"approval_other":0.36361159227665746,"relevance":58.550205399125254,"bioassay":402.06063205179885,"modulator":18.008779201301508,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.24897890653704605,"top_K_average_precision":0.5314329883166508,"moa":0.5214986796794936,"commercial":27.379081529581526,"structure":0.8840474721509741,"druggable":0.752763451465793,"approval_other":0.3984914570155971,"relevance":52.71584377734188,"bioassay":335.0582190628554,"modulator":11.363227256176904,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","top_K_overlap":0.24633041377227424,"top_K_average_precision":0.5071854892464974,"moa":0.5267801426824682,"commercial":27.384602453102456,"structure":0.9121808934425478,"druggable":0.8177643834445382,"approval_other":0.3917276273972308,"relevance":61.30286636102969,"bioassay":347.6086946566479,"modulator":12.454499837070108,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.2688893793544956,"top_K_average_precision":0.4727770557728443,"moa":0.5739829129957037,"commercial":27.942165257223397,"structure":0.8970840676039792,"druggable":0.861356453281232,"approval_other":0.427889913675312,"relevance":50.2684183390127,"bioassay":378.8729873801223,"modulator":15.842621298318202,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","top_K_overlap":0.21820094203815132,"top_K_average_precision":0.3883528950959054,"moa":0.46251210725106073,"commercial":26.861565690123836,"structure":0.8216273416273415,"druggable":0.748000858000858,"approval_other":0.39267338767338766,"relevance":48.9810516655317,"bioassay":372.6868161018161,"modulator":13.95813098813099,"valid_fraction":1,"rank":6}],"name":"Target Prediction for Fibrotic Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-inflammatory-immunological-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Inflammatory & Immunological Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 21 diseases: Allergic rhinitis, Alopecia areata, Asthma, Atopic dermatitis, Celiac disease...","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.400217994124374,"top_K_average_precision":0.6843915055736927,"moa":1.2922516106569464,"commercial":385.9396369803932,"structure":0.9056718868813873,"druggable":0.8220273202553694,"approval_other":0.4456126094645954,"relevance":76.31492676144823,"bioassay":285.7204574395141,"modulator":10.594855821879426,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","top_K_overlap":0.3356993392466449,"top_K_average_precision":0.5735009131060542,"moa":1.2591786286723745,"commercial":387.849712102695,"structure":0.8829638506613717,"druggable":0.7629491129733889,"approval_other":0.3674625124693009,"relevance":75.02084043620223,"bioassay":244.5170701669265,"modulator":10.163002891074346,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.35201736688520124,"top_K_average_precision":0.6789156993177313,"moa":1.192822926294399,"commercial":464.1216843057578,"structure":0.8975360600389367,"druggable":0.7449620860293896,"approval_other":0.3728555536758421,"relevance":70.2604852596506,"bioassay":221.3254003629745,"modulator":8.04687942831505,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","top_K_overlap":0.3451978327464764,"top_K_average_precision":0.6746966470032435,"moa":1.1754865925674045,"commercial":471.292050804112,"structure":0.9049952060966201,"druggable":0.7705636955233581,"approval_other":0.4076035212335488,"relevance":82.4436815121978,"bioassay":317.4238456801878,"modulator":10.49326682072531,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.3796179740053487,"top_K_average_precision":0.7030076067224039,"moa":1.2778401935808237,"commercial":488.7283297585448,"structure":0.9135020256407843,"druggable":0.8145701936061143,"approval_other":0.4160701828203623,"relevance":76.83292492326375,"bioassay":272.61944786678714,"modulator":11.48145187297953,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","top_K_overlap":0.3384010885644758,"top_K_average_precision":0.5832116817426953,"moa":1.2329736175069155,"commercial":464.64923471254724,"structure":0.817593816944245,"druggable":0.7163501488925862,"approval_other":0.42026829960125023,"relevance":63.86038752709129,"bioassay":284.5124354350687,"modulator":10.4994432592648,"valid_fraction":1,"rank":5}],"name":"Target Prediction for Inflammatory & Immunological Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-mental-behavioral-disorders":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Mental & Behavioral Disorders","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 3 diseases: Bipolar disorder, Major depressive disorder, Schizophrenia","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.3987162774753157,"top_K_average_precision":0.6654924673585216,"moa":1.7384726671447925,"commercial":387.8972464242058,"structure":0.8933782267115601,"druggable":0.7154882154882154,"approval_other":0.4595959595959596,"relevance":52.66717774579751,"bioassay":461.4882154882155,"modulator":9.55780022446689,"valid_fraction":1,"rank":3},{"model":"Claude Opus 4.8, high","top_K_overlap":0.3763661541098517,"top_K_average_precision":0.6635782904219824,"moa":1.7783150412646578,"commercial":419.0806754277948,"structure":0.8842307692307694,"druggable":0.647051282051282,"approval_other":0.33897435897435896,"relevance":56.94288263943931,"bioassay":330.3283333333333,"modulator":5.111025641025641,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.43063117109233695,"top_K_average_precision":0.654061567440201,"moa":1.7906526416447524,"commercial":448.24399308672264,"structure":0.8608305274971942,"druggable":0.5845117845117845,"approval_other":0.2684624017957351,"relevance":58.31915520581229,"bioassay":355.61066217732883,"modulator":3.9791245791245786,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","top_K_overlap":0.3464554794263093,"top_K_average_precision":0.5994023470966651,"moa":1.6211036299968278,"commercial":367.3412220623457,"structure":0.8042773600668337,"druggable":0.5542355889724311,"approval_other":0.291328320802005,"relevance":52.01905245701969,"bioassay":389.2468003341687,"modulator":3.740501253132832,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.3976427498545325,"top_K_average_precision":0.6454210226322766,"moa":1.7746493067928437,"commercial":407.7027097851892,"structure":0.878962703962704,"druggable":0.7540209790209791,"approval_other":0.44551282051282054,"relevance":56.78367105555052,"bioassay":471.5974941724942,"modulator":9.46532634032634,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","top_K_overlap":0.4330338281483754,"top_K_average_precision":0.718158567320049,"moa":1.8478005614218678,"commercial":441.7178541250374,"structure":0.9135449735449734,"druggable":0.6671957671957672,"approval_other":0.433015873015873,"relevance":58.00641886055564,"bioassay":481.1059259259259,"modulator":5.517671957671958,"valid_fraction":1,"rank":1}],"name":"Target Prediction for Mental & Behavioral Disorders","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-neurological-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Neurological Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 8 diseases: Alzheimer's disease, Amyotrophic Lateral Sclerosis, Epilepsy, Huntington's Disease, Multiple sclerosis...","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.30835825747920775,"top_K_average_precision":0.5977913663634447,"moa":0.9389085860216796,"commercial":110.9413104228993,"structure":0.8876237237660712,"druggable":0.6000587078035198,"approval_other":0.2632458880149705,"relevance":49.19298983098905,"bioassay":244.33461298031892,"modulator":5.576532374708338,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","top_K_overlap":0.2720650193946528,"top_K_average_precision":0.5754054963845007,"moa":0.8597774830065303,"commercial":96.2320949710663,"structure":0.8754299690761376,"druggable":0.6024584646591002,"approval_other":0.25550274634798903,"relevance":56.95842728689847,"bioassay":223.1913358068603,"modulator":5.246296508937613,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.2842371274917481,"top_K_average_precision":0.5841495923828978,"moa":0.9229663436644465,"commercial":113.64234483498731,"structure":0.8883695210243375,"druggable":0.6308816406580168,"approval_other":0.2810343708279488,"relevance":57.061919296463756,"bioassay":277.52533193424824,"modulator":6.162669517027315,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","top_K_overlap":0.24583667620090915,"top_K_average_precision":0.6043105491788907,"moa":0.7932094403830302,"commercial":102.49581999540578,"structure":0.9028086975836619,"druggable":0.5603251680464358,"approval_other":0.2374481642076467,"relevance":49.10596450375989,"bioassay":187.7026380536899,"modulator":3.96602709436995,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.32244453444164,"top_K_average_precision":0.6330128768357612,"moa":0.9937865220964461,"commercial":124.07546975314771,"structure":0.830946950774214,"druggable":0.6262628798213963,"approval_other":0.2718656233795187,"relevance":51.095524351104814,"bioassay":251.70791390853907,"modulator":5.777106503311235,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","top_K_overlap":0.29853396608657473,"top_K_average_precision":0.5788481689366353,"moa":0.9037295052929484,"commercial":84.35109783609556,"structure":0.8291537210576284,"druggable":0.60229004364757,"approval_other":0.27469160452624625,"relevance":54.02989775886649,"bioassay":369.9238242624477,"modulator":6.3920925887056805,"valid_fraction":1,"rank":3}],"name":"Target Prediction for Neurological Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-ophthalmic-diseases":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Ophthalmic Diseases","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 7 diseases: Age-related macular degeneration, Corneal dystrophy, Diabetic retinopathy, Glaucoma, Pterygium...","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.3985320099704318,"top_K_average_precision":0.8105312137037691,"moa":0.9100962210593695,"commercial":92.67656365068844,"structure":0.7404908471994096,"druggable":0.6622667251575342,"approval_other":0.4106662517895825,"relevance":38.79960441451586,"bioassay":344.05591009409545,"modulator":13.590355383030166,"valid_fraction":1,"rank":3},{"model":"Claude Opus 4.8, high","top_K_overlap":0.37830997421876084,"top_K_average_precision":0.742661774976665,"moa":0.9220381363934591,"commercial":113.70131925834316,"structure":0.7186487683856104,"druggable":0.5983252322726008,"approval_other":0.358980278980279,"relevance":36.986635738119126,"bioassay":343.18277511961725,"modulator":14.473594465962888,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.3895029270229981,"top_K_average_precision":0.7922314776049154,"moa":0.9433755043467228,"commercial":109.99697460959256,"structure":0.7382747247279268,"druggable":0.6367270523674465,"approval_other":0.37131874693451045,"relevance":35.78777009805076,"bioassay":247.2781154258987,"modulator":4.320843907050803,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","top_K_overlap":0.3674470047365048,"top_K_average_precision":0.6411662179833179,"moa":0.6665915361606524,"commercial":92.24768237214964,"structure":0.6766878376064385,"druggable":0.5558630483605966,"approval_other":0.37424596216685135,"relevance":38.45287026711141,"bioassay":306.55093319832616,"modulator":11.258497882198439,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.42789813611556415,"top_K_average_precision":0.836546763032379,"moa":0.9600141551822288,"commercial":109.99697460959256,"structure":0.7347217355913008,"druggable":0.6407340163861903,"approval_other":0.4054363625792197,"relevance":43.12839286541051,"bioassay":478.1358969808659,"modulator":16.37725393539058,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","top_K_overlap":0.40701346054443455,"top_K_average_precision":0.7613228432608768,"moa":0.9781250082426917,"commercial":138.75978586429923,"structure":0.715959965959966,"druggable":0.6138306138306139,"approval_other":0.3463751856609,"relevance":36.73653346474648,"bioassay":372.0978512228512,"modulator":14.64584621727479,"valid_fraction":1,"rank":2}],"name":"Target Prediction for Ophthalmic Diseases","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"dis-targetbench-by-disease-group-target-prediction-for-reproductive-health-conditions":{"meta":{"title":"TargetBench by Disease Group: Target Prediction for Reproductive Health Conditions","category":"Disease Biology > TargetBench by Disease Group","description":"Evaluated on 3 diseases: Endometriosis, Polycystic ovary syndrome, Preeclampsia","metric_types":[],"paper_link":"#","source_file":"targetbench_ci.json","original_category":"Biology"},"radar_keys":["top_K_overlap","top_K_average_precision","moa","commercial","structure","druggable","approval_other","relevance","bioassay","modulator","valid_fraction"],"leaderboard_metric":"top_K_overlap","results":[{"model":"GPT 5.5, medium","top_K_overlap":0.20959595959595959,"top_K_average_precision":0.42286991424922454,"moa":0.7649494949494949,"commercial":48.13080808080807,"structure":1,"druggable":0.9583333333333334,"approval_other":0.6735347985347985,"relevance":42.06419378833173,"bioassay":429.3942307692308,"modulator":9.618131868131869,"valid_fraction":1,"rank":3},{"model":"Claude Opus 4.8, high","top_K_overlap":0.12626262626262627,"top_K_average_precision":0.3783008658008658,"moa":0.7398585858585859,"commercial":56.52191919191919,"structure":0.9866666666666667,"druggable":0.8650000000000001,"approval_other":0.5616666666666666,"relevance":50.93753479385662,"bioassay":408.2833333333333,"modulator":7.001666666666666,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","top_K_overlap":0.23989898989898992,"top_K_average_precision":0.5101045713545713,"moa":0.8036565656565656,"commercial":48.13080808080807,"structure":0.9565217391304347,"druggable":0.8672360248447205,"approval_other":0.6679606625258799,"relevance":37.661996738708396,"bioassay":434.1785714285714,"modulator":13.886387163561077,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","top_K_overlap":0.20959595959595959,"top_K_average_precision":0.4433738748024463,"moa":0.7649494949494949,"commercial":48.13080808080807,"structure":0.9743589743589745,"druggable":0.8942307692307692,"approval_other":0.6804029304029303,"relevance":40.10527042136239,"bioassay":373.95650183150184,"modulator":11.53937728937729,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","top_K_overlap":0.2297979797979798,"top_K_average_precision":0.5002853936187269,"moa":0.7902020202020202,"commercial":48.13080808080807,"structure":0.9722222222222222,"druggable":0.9583333333333334,"approval_other":0.6547619047619048,"relevance":36.38536856927662,"bioassay":394.0853174603174,"modulator":11.013888888888891,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","top_K_overlap":0.18434343434343434,"top_K_average_precision":0.442221803251215,"moa":0.7786565656565655,"commercial":48.13080808080807,"structure":0.9339774557165862,"druggable":0.7747584541062803,"approval_other":0.5964170692431562,"relevance":29.92679230185479,"bioassay":359.90177133655396,"modulator":11.600644122383253,"valid_fraction":1,"rank":5}],"name":"Target Prediction for Reproductive Health Conditions","metric_info":{"top_K_overlap":{"label":"Precision@K","shortLabel":"P@K","direction":"higher_better"},"top_K_average_precision":{"label":"Average Precision@K","shortLabel":"AP@K","direction":"higher_better"},"moa":{"label":"MoA Score","shortLabel":"MoA","direction":"higher_better"},"commercial":{"label":"Commercial Score","shortLabel":"Comm","direction":"higher_better"},"structure":{"label":"3D Structure Available","shortLabel":"Str","direction":"higher_better"},"druggable":{"label":"Druggable Targets","shortLabel":"Drug","direction":"higher_better"},"approval_other":{"label":"Approved Other Indication","shortLabel":"Appr","direction":"higher_better"},"relevance":{"label":"Pathway Relevance","shortLabel":"Rel","direction":"higher_better"},"bioassay":{"label":"Bioassay Count","shortLabel":"Bio","direction":"higher_better"},"modulator":{"label":"Known Modulators","shortLabel":"Mod","direction":"higher_better"},"valid_fraction":{"label":"Valid Fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"top_K_overlap","min":0,"max":1}},"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-all-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: CrossDocked (All-Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"crossdocked_all_conditions_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","anchor_points_3d_passed","opt_anchor_points_3d_passed","pharmacophore_passed","opt_pharmacophore_passed","pocket_ligand_interaction_passed","opt_pocket_ligand_interaction_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":2.98,"opt_unidock_score":-6.294,"posebusters_intermolecular_passed":0.2626582278481013,"opt_posebusters_intermolecular_passed":0.990506329113924,"anchor_points_3d_passed":0.9926160337552743,"opt_anchor_points_3d_passed":0.012658227848101266,"pharmacophore_passed":0.9936708860759493,"opt_pharmacophore_passed":0.9936708860759493,"pocket_ligand_interaction_passed":0.7267932489451476,"opt_pocket_ligand_interaction_passed":0.2795358649789029,"valid_mol_passed":0.9936708860759493,"descriptors_ro5_passed":0.9778481012658228,"descriptors_nheavy":20.92887473460722,"posebusters_intramolecular_passed":0.40717299578059074,"opt_posebusters_intramolecular_passed":0.47890295358649787,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":11.303,"opt_unidock_score":-6.099,"posebusters_intermolecular_passed":0.09071729957805907,"opt_posebusters_intermolecular_passed":0.6993670886075949,"anchor_points_3d_passed":0.7046413502109705,"opt_anchor_points_3d_passed":0.004219409282700422,"pharmacophore_passed":0.7046413502109705,"opt_pharmacophore_passed":0.7046413502109705,"pocket_ligand_interaction_passed":0.43776371308016876,"opt_pocket_ligand_interaction_passed":0.1962025316455696,"valid_mol_passed":0.7046413502109705,"descriptors_ro5_passed":0.6856540084388185,"descriptors_nheavy":20.47005988023952,"posebusters_intramolecular_passed":0.24156118143459915,"opt_posebusters_intramolecular_passed":0.26054852320675104,"rank":2},{"model":"Gemini 3.1 Pro, high","unidock_score":28.990499999999997,"opt_unidock_score":-6.088,"posebusters_intermolecular_passed":0.04746835443037975,"opt_posebusters_intermolecular_passed":0.9219409282700421,"anchor_points_3d_passed":0.9261603375527426,"opt_anchor_points_3d_passed":0.010548523206751054,"pharmacophore_passed":0.929324894514768,"opt_pharmacophore_passed":0.929324894514768,"pocket_ligand_interaction_passed":0.679324894514768,"opt_pocket_ligand_interaction_passed":0.2668776371308017,"valid_mol_passed":0.9314345991561181,"descriptors_ro5_passed":0.79957805907173,"descriptors_nheavy":22.787089467723668,"posebusters_intramolecular_passed":0.3459915611814346,"opt_posebusters_intramolecular_passed":0.38924050632911394,"rank":3}],"name":"CrossDocked (All-Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Raw Conformations)","shortLabel":"Anch3D (Raw)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on raw conformations.","units":"%"},"opt_anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Optimized Conformations)","shortLabel":"Anch3D (Opt)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on optimized conformations.","units":"%"},"pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Raw Conformations)","shortLabel":"Ph4 (Raw)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on raw conformations.","units":"%"},"opt_pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Optimized Conformations)","shortLabel":"Ph4 (Opt)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on optimized conformations.","units":"%"},"pocket_ligand_interaction_passed":{"label":"Success Rate: PLI Points (Raw Conformations)","shortLabel":"PLI (Raw)","direction":"higher_better","description":"Success Rate on per-generation condition Pocket-Ligand Interaction Points filter on raw conformations.","units":"%"},"opt_pocket_ligand_interaction_passed":{"label":"Success Rate: PLI Points (Optimized Conformations)","shortLabel":"PLI (Opt)","direction":"higher_better","description":"Success Rate on per-generation condition Pocket-Ligand Interaction Points filter on optimized conformations.","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-only-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: CrossDocked (Pocket-Only Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"crossdocked_pocket_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":18.794,"opt_unidock_score":-6.544,"posebusters_intermolecular_passed":0.060126582278481014,"opt_posebusters_intermolecular_passed":0.9915611814345991,"valid_mol_passed":0.9926160337552743,"descriptors_ro5_passed":0.990506329113924,"descriptors_nheavy":22.77258235919235,"posebusters_intramolecular_passed":0.22679324894514769,"opt_posebusters_intramolecular_passed":0.30063291139240506,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":56.346,"opt_unidock_score":-6.5405,"posebusters_intermolecular_passed":0.011603375527426161,"opt_posebusters_intermolecular_passed":0.8987341772151899,"valid_mol_passed":0.9008438818565401,"descriptors_ro5_passed":0.8945147679324894,"descriptors_nheavy":21.32903981264637,"posebusters_intramolecular_passed":0.47468354430379744,"opt_posebusters_intramolecular_passed":0.5464135021097046,"rank":2},{"model":"Gemini 3.1 Pro, high","unidock_score":57.494,"opt_unidock_score":-6.433,"posebusters_intermolecular_passed":0.014767932489451477,"opt_posebusters_intermolecular_passed":0.9345991561181435,"valid_mol_passed":0.939873417721519,"descriptors_ro5_passed":0.8533755274261603,"descriptors_nheavy":23.23793490460157,"posebusters_intramolecular_passed":0.410337552742616,"opt_posebusters_intramolecular_passed":0.4883966244725738,"rank":3}],"name":"CrossDocked (Pocket-Only Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-fragment-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: CrossDocked (Pocket-Fragment Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"crossdocked_pocket_fragment_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","anchor_points_3d_passed","opt_anchor_points_3d_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":7.135,"opt_unidock_score":-6.4165,"posebusters_intermolecular_passed":0.2109704641350211,"opt_posebusters_intermolecular_passed":0.9926160337552743,"anchor_points_3d_passed":0.9936708860759493,"opt_anchor_points_3d_passed":0.004219409282700422,"valid_mol_passed":0.9936708860759493,"descriptors_ro5_passed":0.9936708860759493,"descriptors_nheavy":21.55732484076433,"posebusters_intramolecular_passed":0.4578059071729958,"opt_posebusters_intramolecular_passed":0.5516877637130801,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":28.126,"opt_unidock_score":-6.167999999999999,"posebusters_intermolecular_passed":0.06856540084388185,"opt_posebusters_intermolecular_passed":0.8924050632911392,"anchor_points_3d_passed":0.8966244725738397,"opt_anchor_points_3d_passed":0.006329113924050633,"valid_mol_passed":0.8966244725738397,"descriptors_ro5_passed":0.8955696202531646,"descriptors_nheavy":20.554117647058824,"posebusters_intramolecular_passed":0.49683544303797467,"opt_posebusters_intramolecular_passed":0.5432489451476793,"rank":2},{"model":"Gemini 3.1 Pro, high","unidock_score":39.094,"opt_unidock_score":-5.887,"posebusters_intermolecular_passed":0.07278481012658228,"opt_posebusters_intermolecular_passed":0.9440928270042194,"anchor_points_3d_passed":0.9377637130801688,"opt_anchor_points_3d_passed":0.010548523206751054,"valid_mol_passed":0.9493670886075949,"descriptors_ro5_passed":0.8185654008438819,"descriptors_nheavy":22.29222222222222,"posebusters_intramolecular_passed":0.5675105485232067,"opt_posebusters_intramolecular_passed":0.6371308016877637,"rank":3}],"name":"CrossDocked (Pocket-Fragment Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Raw Conformations)","shortLabel":"Anch3D (Raw)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on raw conformations.","units":"%"},"opt_anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Optimized Conformations)","shortLabel":"Anch3D (Opt)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on optimized conformations.","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-pharmacophore-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: CrossDocked (Pocket-Pharmacophore Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"crossdocked_pocket_pharmacophore_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","pharmacophore_passed","opt_pharmacophore_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":12.404499999999999,"opt_unidock_score":-6.387,"posebusters_intermolecular_passed":0.10443037974683544,"opt_posebusters_intermolecular_passed":0.9862869198312236,"pharmacophore_passed":0.990506329113924,"opt_pharmacophore_passed":0.990506329113924,"valid_mol_passed":0.990506329113924,"descriptors_ro5_passed":0.990506329113924,"descriptors_nheavy":21.71352502662407,"posebusters_intramolecular_passed":0.3217299578059072,"opt_posebusters_intramolecular_passed":0.4061181434599156,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":52.326,"opt_unidock_score":-6.4825,"posebusters_intermolecular_passed":0.02109704641350211,"opt_posebusters_intermolecular_passed":0.8449367088607594,"pharmacophore_passed":0.8459915611814346,"opt_pharmacophore_passed":0.8459915611814346,"valid_mol_passed":0.8459915611814346,"descriptors_ro5_passed":0.8459915611814346,"descriptors_nheavy":20.90773067331671,"posebusters_intramolecular_passed":0.41877637130801687,"opt_posebusters_intramolecular_passed":0.4757383966244726,"rank":3},{"model":"Gemini 3.1 Pro, high","unidock_score":47.751,"opt_unidock_score":-6.139,"posebusters_intermolecular_passed":0.0379746835443038,"opt_posebusters_intermolecular_passed":0.9409282700421941,"pharmacophore_passed":0.9419831223628692,"opt_pharmacophore_passed":0.9419831223628692,"valid_mol_passed":0.9462025316455697,"descriptors_ro5_passed":0.8533755274261603,"descriptors_nheavy":23.190635451505017,"posebusters_intramolecular_passed":0.4092827004219409,"opt_posebusters_intramolecular_passed":0.5052742616033755,"rank":2}],"name":"CrossDocked (Pocket-Pharmacophore Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Raw Conformations)","shortLabel":"Ph4 (Raw)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on raw conformations.","units":"%"},"opt_pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Optimized Conformations)","shortLabel":"Ph4 (Opt)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on optimized conformations.","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-all-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: PLInder (All-Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"plinder_all_conditions_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","anchor_points_3d_passed","opt_anchor_points_3d_passed","pharmacophore_passed","opt_pharmacophore_passed","pocket_ligand_interaction_passed","opt_pocket_ligand_interaction_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":1.653,"opt_unidock_score":-5.7,"posebusters_intermolecular_passed":0.36633663366336633,"opt_posebusters_intermolecular_passed":0.9900990099009901,"anchor_points_3d_passed":0.9861386138613861,"opt_anchor_points_3d_passed":0.0039603960396039604,"pharmacophore_passed":0.9920792079207921,"opt_pharmacophore_passed":0.9900990099009901,"pocket_ligand_interaction_passed":0.8158415841584158,"opt_pocket_ligand_interaction_passed":0.27524752475247527,"valid_mol_passed":0.9920792079207921,"descriptors_ro5_passed":0.9801980198019802,"descriptors_nheavy":21.389221556886227,"posebusters_intramolecular_passed":0.4198019801980198,"opt_posebusters_intramolecular_passed":0.4792079207920792,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":6.42,"opt_unidock_score":-5.502000000000001,"posebusters_intermolecular_passed":0.1485148514851485,"opt_posebusters_intermolecular_passed":0.7603960396039604,"anchor_points_3d_passed":0.7623762376237624,"opt_anchor_points_3d_passed":0.0039603960396039604,"pharmacophore_passed":0.7623762376237624,"opt_pharmacophore_passed":0.7603960396039604,"pocket_ligand_interaction_passed":0.5247524752475248,"opt_pocket_ligand_interaction_passed":0.2198019801980198,"valid_mol_passed":0.7623762376237624,"descriptors_ro5_passed":0.7405940594059406,"descriptors_nheavy":20.683116883116885,"posebusters_intramolecular_passed":0.32475247524752476,"opt_posebusters_intramolecular_passed":0.3504950495049505,"rank":2},{"model":"Gemini 3.1 Pro, high","unidock_score":16.534,"opt_unidock_score":-5.463,"posebusters_intermolecular_passed":0.08514851485148515,"opt_posebusters_intermolecular_passed":0.9306930693069307,"anchor_points_3d_passed":0.9445544554455445,"opt_anchor_points_3d_passed":0.005940594059405941,"pharmacophore_passed":0.9445544554455445,"opt_pharmacophore_passed":0.9425742574257425,"pocket_ligand_interaction_passed":0.7287128712871287,"opt_pocket_ligand_interaction_passed":0.25544554455445545,"valid_mol_passed":0.9504950495049505,"descriptors_ro5_passed":0.8435643564356435,"descriptors_nheavy":23.185416666666665,"posebusters_intramolecular_passed":0.3188118811881188,"opt_posebusters_intramolecular_passed":0.37623762376237624,"rank":3}],"name":"PLInder (All-Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Raw Conformations)","shortLabel":"Anch3D (Raw)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on raw conformations.","units":"%"},"opt_anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Optimized Conformations)","shortLabel":"Anch3D (Opt)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on optimized conformations.","units":"%"},"pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Raw Conformations)","shortLabel":"Ph4 (Raw)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on raw conformations.","units":"%"},"opt_pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Optimized Conformations)","shortLabel":"Ph4 (Opt)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on optimized conformations.","units":"%"},"pocket_ligand_interaction_passed":{"label":"Success Rate: PLI Points (Raw Conformations)","shortLabel":"PLI (Raw)","direction":"higher_better","description":"Success Rate on per-generation condition Pocket-Ligand Interaction Points filter on raw conformations.","units":"%"},"opt_pocket_ligand_interaction_passed":{"label":"Success Rate: PLI Points (Optimized Conformations)","shortLabel":"PLI (Opt)","direction":"higher_better","description":"Success Rate on per-generation condition Pocket-Ligand Interaction Points filter on optimized conformations.","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-only-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: PLInder (Pocket-Only Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"plinder_pocket_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":19.8675,"opt_unidock_score":-6.099,"posebusters_intermolecular_passed":0.09306930693069307,"opt_posebusters_intermolecular_passed":0.996039603960396,"valid_mol_passed":0.998019801980198,"descriptors_ro5_passed":0.998019801980198,"descriptors_nheavy":22.648809523809526,"posebusters_intramolecular_passed":0.3108910891089109,"opt_posebusters_intramolecular_passed":0.37425742574257426,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":53.141000000000005,"opt_unidock_score":-6.208500000000001,"posebusters_intermolecular_passed":0.03762376237623762,"opt_posebusters_intermolecular_passed":0.9524752475247524,"valid_mol_passed":0.9564356435643564,"descriptors_ro5_passed":0.9544554455445544,"descriptors_nheavy":21.306418219461698,"posebusters_intramolecular_passed":0.5287128712871287,"opt_posebusters_intramolecular_passed":0.5801980198019802,"rank":3},{"model":"Gemini 3.1 Pro, high","unidock_score":51.342,"opt_unidock_score":-5.9559999999999995,"posebusters_intermolecular_passed":0.03564356435643564,"opt_posebusters_intermolecular_passed":0.9445544554455445,"valid_mol_passed":0.9504950495049505,"descriptors_ro5_passed":0.8950495049504951,"descriptors_nheavy":23.216666666666665,"posebusters_intramolecular_passed":0.41188118811881186,"opt_posebusters_intramolecular_passed":0.4772277227722772,"rank":2}],"name":"PLInder (Pocket-Only Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-fragment-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: PLInder (Pocket-Fragment Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"plinder_pocket_fragment_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","anchor_points_3d_passed","opt_anchor_points_3d_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":2.9335,"opt_unidock_score":-5.9115,"posebusters_intermolecular_passed":0.3188118811881188,"opt_posebusters_intermolecular_passed":0.9861386138613861,"anchor_points_3d_passed":0.9881188118811881,"opt_anchor_points_3d_passed":0.005940594059405941,"valid_mol_passed":0.9881188118811881,"descriptors_ro5_passed":0.9861386138613861,"descriptors_nheavy":21.803607214428858,"posebusters_intramolecular_passed":0.5267326732673268,"opt_posebusters_intramolecular_passed":0.592079207920792,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":22.5275,"opt_unidock_score":-5.737,"posebusters_intermolecular_passed":0.12673267326732673,"opt_posebusters_intermolecular_passed":0.9485148514851485,"anchor_points_3d_passed":0.9485148514851485,"opt_anchor_points_3d_passed":0.0019801980198019802,"valid_mol_passed":0.9504950495049505,"descriptors_ro5_passed":0.9485148514851485,"descriptors_nheavy":20.541666666666668,"posebusters_intramolecular_passed":0.6277227722772277,"opt_posebusters_intramolecular_passed":0.6673267326732674,"rank":3},{"model":"Gemini 3.1 Pro, high","unidock_score":21.182,"opt_unidock_score":-5.373,"posebusters_intermolecular_passed":0.11287128712871287,"opt_posebusters_intermolecular_passed":0.9564356435643564,"anchor_points_3d_passed":0.9524752475247524,"opt_anchor_points_3d_passed":0.0019801980198019802,"valid_mol_passed":0.9584158415841584,"descriptors_ro5_passed":0.8752475247524752,"descriptors_nheavy":22.450413223140497,"posebusters_intramolecular_passed":0.5722772277227722,"opt_posebusters_intramolecular_passed":0.6574257425742575,"rank":2}],"name":"PLInder (Pocket-Fragment Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Raw Conformations)","shortLabel":"Anch3D (Raw)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on raw conformations.","units":"%"},"opt_anchor_points_3d_passed":{"label":"Success Rate: Anchor Points 3D (Optimized Conformations)","shortLabel":"Anch3D (Opt)","direction":"higher_better","description":"Success Rate on Anchor Points 3D filter on optimized conformations.","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-pharmacophore-condition":{"meta":{"title":"Bench3DFit: Molecular Generation with 3D Conditions: PLInder (Pocket-Pharmacophore Condition)","category":"Medicinal Chemistry > Bench3DFit: Molecular Generation with 3D Conditions","description":"plinder_pocket_pharmacophore_sdf_enum","metric_types":[],"paper_link":"#","source_file":"bench3dfit_aggregated_only_general_purpose_llms.json","original_category":"Molecular Design & Optimization"},"radar_keys":["unidock_score","opt_unidock_score","posebusters_intermolecular_passed","opt_posebusters_intermolecular_passed","pharmacophore_passed","opt_pharmacophore_passed","valid_mol_passed","descriptors_ro5_passed","descriptors_nheavy","posebusters_intramolecular_passed","opt_posebusters_intramolecular_passed"],"leaderboard_metric":"unidock_score","results":[{"model":"GPT 5.5, medium","unidock_score":11.029,"opt_unidock_score":-5.9395,"posebusters_intermolecular_passed":0.1603960396039604,"opt_posebusters_intermolecular_passed":0.9920792079207921,"pharmacophore_passed":0.996039603960396,"opt_pharmacophore_passed":0.994059405940594,"valid_mol_passed":0.996039603960396,"descriptors_ro5_passed":0.9920792079207921,"descriptors_nheavy":22.15506958250497,"posebusters_intramolecular_passed":0.29108910891089107,"opt_posebusters_intramolecular_passed":0.38811881188118813,"rank":1},{"model":"Claude Opus 4.8, high","unidock_score":40.8,"opt_unidock_score":-6.064500000000001,"posebusters_intermolecular_passed":0.04950495049504951,"opt_posebusters_intermolecular_passed":0.8950495049504951,"pharmacophore_passed":0.897029702970297,"opt_pharmacophore_passed":0.8950495049504951,"valid_mol_passed":0.897029702970297,"descriptors_ro5_passed":0.897029702970297,"descriptors_nheavy":20.799116997792495,"posebusters_intramolecular_passed":0.4396039603960396,"opt_posebusters_intramolecular_passed":0.497029702970297,"rank":3},{"model":"Gemini 3.1 Pro, high","unidock_score":36.7605,"opt_unidock_score":-5.651,"posebusters_intermolecular_passed":0.08712871287128712,"opt_posebusters_intermolecular_passed":0.9405940594059405,"pharmacophore_passed":0.9465346534653465,"opt_pharmacophore_passed":0.9425742574257425,"valid_mol_passed":0.9465346534653465,"descriptors_ro5_passed":0.8613861386138614,"descriptors_nheavy":23.09623430962343,"posebusters_intramolecular_passed":0.42574257425742573,"opt_posebusters_intramolecular_passed":0.5168316831683168,"rank":2}],"name":"PLInder (Pocket-Pharmacophore Condition)","metric_info":{"unidock_score":{"label":"Unidock Score (Raw Conformations)","shortLabel":"Unidock (Raw)","direction":"lower_better","description":"Unidock score of raw conformations, averaged.","units":"kcal/mol"},"opt_unidock_score":{"label":"Unidock Score (Optimized Conformations)","shortLabel":"Unidock (Opt)","direction":"lower_better","description":"Unidock score of optimized conformations, averaged.","units":"kcal/mol"},"posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Raw Conformations)","shortLabel":"PB Inter. (Raw)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on raw conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"opt_posebusters_intermolecular_passed":{"label":"Success Rate: PoseBusters Intermolecular (Optimized Conformations)","shortLabel":"PB Inter. (Opt)","direction":"higher_better","description":"Success Rate on 3 PoseBusters intermolecular (Interaction Quality) filters on optimized conformations: Protein-Ligand Maximum Distance, Minimum Distance to Protein, Volume Overlap with Protein","units":"%"},"pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Raw Conformations)","shortLabel":"Ph4 (Raw)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on raw conformations.","units":"%"},"opt_pharmacophore_passed":{"label":"Success Rate: Pharmacophore Points (Optimized Conformations)","shortLabel":"Ph4 (Opt)","direction":"higher_better","description":"Success Rate on per-generation condition Pharmacophore Points filter on optimized conformations.","units":"%"},"valid_mol_passed":{"label":"valid_mol_passed","shortLabel":"VALID","direction":"higher_better"},"descriptors_ro5_passed":{"label":"descriptors_ro5_passed","shortLabel":"DESCR","direction":"higher_better"},"descriptors_nheavy":{"label":"descriptors_nheavy","shortLabel":"DESCR","direction":"higher_better"},"posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Raw Conformations)","shortLabel":"PB Intra. (Raw)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on raw conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"},"opt_posebusters_intramolecular_passed":{"label":"Success Rate: PoseBusters Intramolecular (Optimized Conformations)","shortLabel":"PB Intra. (Opt)","direction":"higher_better","description":"Success Rate on 6 PoseBusters intramolecular (Pose Quality) filters on optimized conformations: Bond Lengths, Bond Angles, Internal Steric Clash, Aromatic Ring Flatness, Non-Aromatic Ring Non-Flatness, Double Bond Flatness","units":"%"}}},"med-ism-benchmarks-admet-logp":{"meta":{"title":"ISM Benchmarks: ADMET: LogP","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"LogP (partition coefficient) measures the lipophilicity of a neutral compound between octanol and water. A fundamental physicochemical descriptor influencing membrane permeability and drug-likeness. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.44368900657414173,"r2_score":0.8331611530394201,"spearman":0.9245431508757529,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.5443313245675675,"r2_score":0.783567323081922,"spearman":0.9067401764084668,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.5292299123447772,"r2_score":0.8035179936094562,"spearman":0.90978250478973,"validity":1,"rank":5},{"model":"Gemini 3.1 Pro, high","mae":0.45180275229357797,"r2_score":0.7947347201516303,"spearman":0.9182982075687579,"validity":0.9952520087655223,"rank":3},{"model":"Kimi K3","mae":0.7112597967178546,"r2_score":0.683944064831105,"spearman":0.8252905809792114,"validity":0.9941563184806428,"rank":8},{"model":"Grok 4.5, high","mae":0.505340211833455,"r2_score":0.8072011901178524,"spearman":0.9099011757546152,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":0.42203148215783703,"spearman":0.9341017044508095,"r2_score":0.8454417732290027,"validity":0.9996347699050402,"rank":1},{"model":"GPT 5.6 Sol, high","mae":0.5445618682257947,"spearman":0.9039029493128942,"r2_score":0.7797666327895659,"validity":0.9996347699050402,"rank":7}],"name":"LogP","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-logd":{"meta":{"title":"ISM Benchmarks: ADMET: LogD","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"LogD (distribution coefficient at pH 7.4) measures the ratio of a compound's concentration in octanol vs aqueous phase accounting for ionization. A key physicochemical property affecting absorption and distribution. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7539980220195931,"r2_score":0.392486622949316,"spearman":0.6968937323478737,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.8198466961168268,"r2_score":0.2967262734182551,"spearman":0.6579428755334624,"validity":0.9998116051243406,"rank":4},{"model":"Claude Opus 4.8, high","mae":0.764825735282811,"r2_score":0.3919462332872523,"spearman":0.6321284071426769,"validity":0.9918990203466466,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":0.8224964617939233,"r2_score":0.23708630863516977,"spearman":0.6911761073379067,"validity":0.9983044461190655,"rank":5},{"model":"Kimi K3","mae":1.0017884431908484,"r2_score":-0.021558848514310558,"spearman":0.46157593251449147,"validity":0.9881311228334589,"rank":8},{"model":"Grok 4.5, high","mae":0.828575503686511,"r2_score":0.31722575854523827,"spearman":0.5612116515749241,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":0.691626441927542,"spearman":0.7198503424830698,"r2_score":0.5003604758620811,"validity":0.9412207987942728,"rank":1},{"model":"GPT 5.6 Sol, high","mae":0.8289196757807084,"spearman":0.6686422925218468,"r2_score":0.284854160313351,"validity":1,"rank":7}],"name":"LogD","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-kinetic-solubility":{"meta":{"title":"ISM Benchmarks: ADMET: Kinetic Solubility","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Kinetic solubility measures the concentration at which a compound precipitates from a DMSO stock solution in aqueous buffer. A critical early-stage screening parameter for compound prioritization in drug discovery. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7619014500883393,"r2_score":-0.11446407869251174,"spearman":0.31407314332939046,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":1.0270671157755193,"r2_score":-0.8679984965199787,"spearman":0.24634222642770867,"validity":0.9995583038869258,"rank":8},{"model":"Claude Opus 4.8, high","mae":0.753489756564308,"r2_score":-0.06283742632924194,"spearman":0.2635282028062741,"validity":0.9924911660777385,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.868362354268023,"r2_score":-0.474814081445732,"spearman":0.30377066104591605,"validity":0.9986749116607774,"rank":5},{"model":"Kimi K3","mae":0.9520305271836007,"r2_score":-0.6452785716145535,"spearman":0.15343676908514617,"validity":0.991166077738516,"rank":6},{"model":"Grok 4.5, high","mae":0.8011342857773852,"r2_score":-0.1376175759775149,"spearman":0.1457697080481661,"validity":1,"rank":3},{"model":"Claude Opus 5, high","mae":0.8133271543953593,"spearman":0.322880261484678,"r2_score":-0.25703770872481435,"validity":0.9898409893992933,"rank":4},{"model":"GPT 5.6 Sol, high","mae":1.0046223069787985,"spearman":0.26097615853680567,"r2_score":-0.7996880919685556,"validity":1,"rank":7}],"name":"Kinetic Solubility","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-thermodynamic-solubility":{"meta":{"title":"ISM Benchmarks: ADMET: Thermodynamic Solubility","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Thermodynamic solubility is the equilibrium solubility of the most stable crystalline form in aqueous media. A definitive measure of compound solubility important for formulation development. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8417887701367928,"r2_score":0.5363499728289425,"spearman":0.6960108618260085,"validity":1,"rank":3},{"model":"GPT 5.6-sol","mae":0.9193546074509831,"r2_score":0.4649178746088928,"spearman":0.6775777839930748,"validity":0.999751552795031,"rank":5},{"model":"Claude Opus 4.8, high","mae":0.8355916222160226,"r2_score":0.5600747793933172,"spearman":0.6794688037498563,"validity":0.9908074534161491,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":1.0033618415303571,"r2_score":0.3116984321769557,"spearman":0.6519797069681282,"validity":1,"rank":7},{"model":"Kimi K3","mae":1.060656455794326,"r2_score":0.3124278492132454,"spearman":0.5662353624059983,"validity":0.9838509316770186,"rank":8},{"model":"Grok 4.5, high","mae":0.9614420542963058,"r2_score":0.44178755566376726,"spearman":0.576438447805188,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":0.7579712189367719,"spearman":0.7523394018162164,"r2_score":0.632199595462023,"validity":0.9773913043478261,"rank":1},{"model":"GPT 5.6 Sol, high","mae":0.901366097178344,"spearman":0.6814676335707603,"r2_score":0.48104314956048744,"validity":0.9990062111801242,"rank":4}],"name":"Thermodynamic Solubility","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-caco-2-permeability":{"meta":{"title":"ISM Benchmarks: ADMET: Caco-2 Permeability","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Caco-2 cell effective permeability (Papp) using human colon carcinoma cells to model passive diffusion and active transport in the intestine. A key predictor of oral drug absorption. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5940018120785244,"r2_score":0.0969423572466106,"spearman":0.45817510727455685,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.7191803801921604,"r2_score":-0.3603294137831172,"spearman":0.384937072770897,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":0.5970860380705373,"r2_score":0.11318158533928957,"spearman":0.46391709724648555,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":0.7261665293002335,"r2_score":-0.359039753293934,"spearman":0.4105742532165223,"validity":0.9969293756397134,"rank":7},{"model":"Kimi K3","mae":0.7536066908281865,"r2_score":-0.5993576328626371,"spearman":0.26485775371122894,"validity":0.9938587512794268,"rank":8},{"model":"Grok 4.5, high","mae":0.692096265503074,"r2_score":-0.22983961641061046,"spearman":0.24873814232028582,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":0.5950572193784825,"spearman":0.49337757966423745,"r2_score":0.09017644539409542,"validity":0.9401228249744115,"rank":2},{"model":"GPT 5.6 Sol, high","mae":0.7247393827257601,"spearman":0.39554594830657935,"r2_score":-0.35571742981624377,"validity":1,"rank":6}],"name":"Caco-2 Permeability","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-caco-2-efflux-ratio":{"meta":{"title":"ISM Benchmarks: ADMET: Caco-2 Efflux Ratio","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Caco-2 efflux ratio measures the ratio of basolateral-to-apical vs apical-to-basolateral transport, indicating active efflux by transporters like P-gp. High efflux ratios suggest limited oral absorption. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5406448820284098,"r2_score":0.1292650221965964,"spearman":0.36856731371516094,"validity":1,"rank":3},{"model":"GPT 5.6-sol","mae":0.5631872382385745,"r2_score":0.008291737794085985,"spearman":0.35054573799153665,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.5332258882049226,"r2_score":0.14936378011846507,"spearman":0.42094680114548283,"validity":0.982982982982983,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":0.5471486996321026,"r2_score":0.047889000230960654,"spearman":0.3417855707759149,"validity":1,"rank":4},{"model":"Kimi K3","mae":0.6964023409830298,"r2_score":-0.5176661955848885,"spearman":0.13258143189321997,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.6058765452851187,"r2_score":-0.11600158391027193,"spearman":0.11178488632391108,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.5329561121022741,"spearman":0.4043804956607306,"r2_score":0.11356311042281053,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mae":0.5566647954993521,"spearman":0.38723284937103847,"r2_score":0.037962762565731056,"validity":1,"rank":5}],"name":"Caco-2 Efflux Ratio","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-pampa-permeability":{"meta":{"title":"ISM Benchmarks: ADMET: PAMPA Permeability","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Parallel Artificial Membrane Permeability Assay (PAMPA) provides a cell-free, high-throughput measure of passive transcellular permeability. Complementary to Caco-2 for absorption prediction. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8889248352820773,"r2_score":0.15450640784885528,"spearman":0.4552374895047629,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":1.124212094960285,"r2_score":-0.3263851768987287,"spearman":0.3445117697456384,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":0.9472682702665975,"r2_score":0.07019684869636189,"spearman":0.4275920596947771,"validity":0.9816700610997964,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":1.5834450877420352,"r2_score":-1.657100564015321,"spearman":0.3901364027754415,"validity":0.9908350305498982,"rank":8},{"model":"Kimi K3","mae":1.1959642227342158,"r2_score":-0.5500875898507465,"spearman":0.2703985425604228,"validity":1,"rank":7},{"model":"Grok 4.5, high","mae":1.032894683404277,"r2_score":-0.18444329601915022,"spearman":0.24231535989775252,"validity":1,"rank":3},{"model":"Claude Opus 5, high","mae":1.115212289751534,"spearman":0.412980092857146,"r2_score":-0.314180030489672,"validity":0.9959266802443992,"rank":4},{"model":"GPT 5.6 Sol, high","mae":1.1326558130621183,"spearman":0.3395789305051922,"r2_score":-0.3571199187616929,"validity":1,"rank":6}],"name":"PAMPA Permeability","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-mdck-mdr1-permeability":{"meta":{"title":"ISM Benchmarks: ADMET: MDCK-MDR1 Permeability","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"MDCK-MDR1 apparent permeability (Papp) measures drug transport across MDCK cells expressing the MDR1 P-glycoprotein transporter. Used to assess intestinal absorption and blood-brain barrier penetration. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5418640547788361,"r2_score":0.08868378875275895,"spearman":0.43858412756906917,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.8422823794969069,"r2_score":-1.8143163220698018,"spearman":0.37774964754386436,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.5878166611468598,"r2_score":0.07876924182903777,"spearman":0.5326799423505876,"validity":0.9768211920529801,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":0.714244000940548,"r2_score":-0.5221651393333109,"spearman":0.31695081270647996,"validity":0.9966887417218543,"rank":4},{"model":"Kimi K3","mae":1.16477099444784,"r2_score":-12.703157485605248,"spearman":0.22318384819657466,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.6814053424279118,"r2_score":-0.34746634199397786,"spearman":0.08594634213821775,"validity":1,"rank":3},{"model":"Claude Opus 5, high","mae":0.9697766786762746,"spearman":0.4681188224582966,"r2_score":-4.653567297461613,"validity":1,"rank":7},{"model":"GPT 5.6 Sol, high","mae":0.7982916055917747,"spearman":0.3606515058096295,"r2_score":-1.1059906287842662,"validity":1,"rank":5}],"name":"MDCK-MDR1 Permeability","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-mdck-mdr1-efflux-ratio":{"meta":{"title":"ISM Benchmarks: ADMET: MDCK-MDR1 Efflux Ratio","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"MDCK-MDR1 efflux ratio measures P-gp-mediated efflux using MDCK cells overexpressing MDR1. High efflux ratios indicate P-gp substrates likely to have limited oral absorption or CNS penetration. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.45531642478098727,"r2_score":0.04940928977932346,"spearman":0.4322011369200442,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.6645122474544432,"r2_score":-2.3777873305362665,"spearman":0.3194311333442939,"validity":1,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.4314743121269036,"r2_score":0.06668966698505263,"spearman":0.5154029949564344,"validity":0.9852941176470589,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.4664361749142044,"r2_score":-0.19662525866538494,"spearman":0.24690045094308197,"validity":0.9957983193277311,"rank":3},{"model":"Kimi K3","mae":0.6044306886191694,"r2_score":-1.100386627463469,"spearman":0.16783707513200094,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.4800766795606904,"r2_score":-0.19931711843290545,"spearman":0.14853697720285503,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":1.5179332570403399,"spearman":0.43125266130837275,"r2_score":-49.245086823367664,"validity":1,"rank":8},{"model":"GPT 5.6 Sol, high","mae":0.5889270568452303,"spearman":0.38444872777201544,"r2_score":-1.678504005652103,"validity":1,"rank":5}],"name":"MDCK-MDR1 Efflux Ratio","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-ppb-human":{"meta":{"title":"ISM Benchmarks: ADMET: PPB Human","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Plasma protein binding (PPB) in human plasma. Only the unbound fraction of a drug can cross membranes and engage its target. Critical for efficacy predictions and dose selection. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.08809912836964284,"r2_score":0.4995210437287827,"spearman":0.6265042784537084,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.09173942598750001,"r2_score":0.4726797927856764,"spearman":0.5864573648577893,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","mae":0.10731725865225226,"r2_score":0.39474300594523215,"spearman":0.6296027358052589,"validity":0.9910714285714286,"rank":6},{"model":"Gemini 3.1 Pro, high","mae":0.09937934265535715,"r2_score":0.3591001387139935,"spearman":0.6174601479593359,"validity":1,"rank":5},{"model":"Kimi K3","mae":0.1377168069375,"r2_score":0.10195382169024025,"spearman":0.4747554742824704,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.10734196170178571,"r2_score":0.32941147239053414,"spearman":0.37878575710177126,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.09560775931964287,"spearman":0.6525531705367292,"r2_score":0.4533726046448049,"validity":1,"rank":4},{"model":"GPT 5.6 Sol, high","mae":0.09217133074821428,"spearman":0.6106932921839018,"r2_score":0.4702522868685205,"validity":1,"rank":3}],"name":"PPB Human","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-ppb-mouse":{"meta":{"title":"ISM Benchmarks: ADMET: PPB Mouse","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Plasma protein binding (PPB) in mouse plasma. Species-specific binding differences are essential for translating preclinical data to human. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.09884695473251029,"r2_score":0.15125741886772004,"spearman":0.4936478693195037,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.10017864197530864,"r2_score":0.1340956434808005,"spearman":0.4294788544530939,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","mae":0.106589625,"r2_score":0.10771311303601783,"spearman":0.4797920109315793,"validity":0.9876543209876543,"rank":6},{"model":"Gemini 3.1 Pro, high","mae":0.1010174074074074,"r2_score":0.1362747043326461,"spearman":0.5082177608183808,"validity":1,"rank":5},{"model":"Kimi K3","mae":0.14634078189300412,"r2_score":-0.2801531019565675,"spearman":0.2506729934763635,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.11465353909465019,"r2_score":-0.0372146491485188,"spearman":0.270677852422952,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.09934539094650205,"spearman":0.5211331651708724,"r2_score":0.17702832868446428,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":0.09845888888888887,"spearman":0.4422244395684094,"r2_score":0.1535215893694205,"validity":1,"rank":1}],"name":"PPB Mouse","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-ppb-rat":{"meta":{"title":"ISM Benchmarks: ADMET: PPB Rat","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Plasma protein binding (PPB) in rat plasma. Species-specific binding is critical for PK translation from the most common preclinical species. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.08232799713544668,"r2_score":0.30595417051599283,"spearman":0.6631247432365347,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.09127280991930836,"r2_score":0.22503741080566597,"spearman":0.6016457936768452,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.08264297764327486,"r2_score":0.36793604278141845,"spearman":0.6815690609873257,"validity":0.9855907780979827,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":0.07630390823188407,"r2_score":0.48069771913788617,"spearman":0.6571765946138783,"validity":0.9942363112391931,"rank":1},{"model":"Kimi K3","mae":0.11385705577521615,"r2_score":0.06428966463556118,"spearman":0.5149064201703465,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.10268938042074928,"r2_score":0.06286180559733179,"spearman":0.3432681313218032,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.08277828536023055,"spearman":0.655363904892335,"r2_score":0.3731957698694849,"validity":1,"rank":4},{"model":"GPT 5.6 Sol, high","mae":0.08810335740634007,"spearman":0.6397696965592115,"r2_score":0.23709948985003415,"validity":1,"rank":5}],"name":"PPB Rat","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-vdss":{"meta":{"title":"ISM Benchmarks: ADMET: VDss","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Volume of distribution at steady state (VDss) measures how extensively a drug distributes from plasma into body tissues. A key PK parameter for dose calculation. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":3.2839506172839505,"r2_score":0.3878252802063542,"spearman":0.9163489950787059,"validity":1,"rank":4},{"model":"GPT 5.6-sol","mae":3.6176543209876546,"r2_score":0.28189774731301764,"spearman":0.8774052205004681,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":3.2800624999999997,"r2_score":0.8771062034479848,"spearman":0.7228663540942227,"validity":0.9876543209876543,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.3212499999999998,"r2_score":0.51015061654804,"spearman":0.9125785264263153,"validity":0.9876543209876543,"rank":1},{"model":"Kimi K3","mae":4.815925925925926,"r2_score":0.250094574297942,"spearman":0.6575749365659252,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":3.8592592592592596,"r2_score":0.3128122714903052,"spearman":0.796005220278536,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":2.3552500000000003,"spearman":0.8992665554529978,"r2_score":0.7957492077664314,"validity":0.9876543209876543,"rank":2},{"model":"GPT 5.6 Sol, high","mae":3.820740740740741,"spearman":0.8899011689934934,"r2_score":0.3406259963402918,"validity":1,"rank":6}],"name":"VDss","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-half-life-human":{"meta":{"title":"ISM Benchmarks: ADMET: Half Life Human","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Human plasma half-life (T1/2) in hours. Determines how long a drug remains active in the body and directly impacts dosing regimen design. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":4.09672891566265,"r2_score":0.049287136849273216,"spearman":0.7003769396828292,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":4.050602409638554,"r2_score":-0.030983450112504363,"spearman":0.7203053328600654,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":4.110962962962963,"r2_score":0.29634006395263957,"spearman":0.7070031423979525,"validity":0.9759036144578314,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":4.508599999999999,"r2_score":-0.5232347644243336,"spearman":0.7465271891372488,"validity":0.9036144578313253,"rank":6},{"model":"Kimi K3","mae":4.398975903614458,"r2_score":0.2878270690651221,"spearman":0.665480540027259,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":4.297831325301205,"r2_score":0.06397145385524139,"spearman":0.5926794701549828,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":4.613253012048192,"spearman":0.6838088099969711,"r2_score":-0.0774689546483085,"validity":1,"rank":7},{"model":"GPT 5.6 Sol, high","mae":4.833759036144579,"spearman":0.6877229049552057,"r2_score":-0.3457767229504476,"validity":1,"rank":8}],"name":"Half Life Human","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-clearance-human-liver-microsomes":{"meta":{"title":"ISM Benchmarks: ADMET: Clearance Human Liver Microsomes","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Intrinsic clearance in human liver microsomes (HLM) evaluates CYP-mediated oxidative metabolism. A standard early-stage assay for predicting hepatic clearance. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.4949594177991138,"r2_score":0.00922843443209631,"spearman":0.23750846856808444,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.5935680226959092,"r2_score":-0.3578658467286866,"spearman":0.16146598165452833,"validity":0.999015263417036,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.4982866804423459,"r2_score":0.002672626178896076,"spearman":0.2476262061343062,"validity":0.9906450024618415,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":0.5157813367110453,"r2_score":-0.0672083810509394,"spearman":0.2524028205628037,"validity":0.9985228951255539,"rank":4},{"model":"Kimi K3","mae":0.6180379409962641,"r2_score":-0.5196823994264124,"spearman":0.08188843497469783,"validity":0.9884293451501723,"rank":8},{"model":"Grok 4.5, high","mae":0.5424363992857142,"r2_score":-0.13946696367157085,"spearman":0.06571417755839232,"validity":0.999507631708518,"rank":5},{"model":"Claude Opus 5, high","mae":0.5142609653523277,"spearman":0.27754627403398563,"r2_score":-0.05285460731090952,"validity":0.9677498769079271,"rank":3},{"model":"GPT 5.6 Sol, high","mae":0.583476812093596,"spearman":0.184853714775576,"r2_score":-0.31470625552202947,"validity":0.999507631708518,"rank":6}],"name":"Clearance Human Liver Microsomes","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-clearance-mouse-liver-microsomes":{"meta":{"title":"ISM Benchmarks: ADMET: Clearance Mouse Liver Microsomes","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Intrinsic clearance in mouse liver microsomes (MLM). Mouse is a common preclinical species; predicting species-specific clearance is essential for PK translation. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5486739389906138,"r2_score":0.03176215639142177,"spearman":0.24285489919085304,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.61379386958237,"r2_score":-0.19281786951818214,"spearman":0.22247399100058624,"validity":0.9992779783393502,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.5886574176357039,"r2_score":-0.07535024041016203,"spearman":0.21054265920531218,"validity":0.9949458483754513,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.5707206937830803,"r2_score":-0.05720596944490719,"spearman":0.2676386573937361,"validity":0.9985559566787003,"rank":2},{"model":"Kimi K3","mae":0.6742101411566787,"r2_score":-0.43048775627876723,"spearman":0.1117762645442146,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.5993725729256317,"r2_score":-0.13005287216505734,"spearman":0.041472693187257985,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.5853723074436417,"spearman":0.24049340448392173,"r2_score":-0.059143192857447735,"validity":0.9992779783393502,"rank":3},{"model":"GPT 5.6 Sol, high","mae":0.6107770379075812,"spearman":0.20952549788930888,"r2_score":-0.1871369621332657,"validity":1,"rank":6}],"name":"Clearance Mouse Liver Microsomes","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-clearance-rat-liver-microsomes":{"meta":{"title":"ISM Benchmarks: ADMET: Clearance Rat Liver Microsomes","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Intrinsic clearance in rat liver microsomes (RLM). Rat is the most widely used preclinical species for PK studies. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.4696905824428827,"r2_score":0.04940163759740657,"spearman":0.30756450454585543,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.5643932253888305,"r2_score":-0.31243723925771283,"spearman":0.22236532232454914,"validity":1,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.4753827453515625,"r2_score":0.042418053390597565,"spearman":0.31048032602222303,"validity":0.9886439545758183,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":0.4868890831463165,"r2_score":-0.034022411260792085,"spearman":0.29154197417776523,"validity":0.9946559786239145,"rank":3},{"model":"Kimi K3","mae":0.6113306543315036,"r2_score":-0.5521697595248305,"spearman":0.1115592343315418,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.5392090497360648,"r2_score":-0.18040738113310706,"spearman":0.04738469193712984,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.4925467856771459,"spearman":0.3048822844588818,"r2_score":-0.021204883034581234,"validity":1,"rank":4},{"model":"GPT 5.6 Sol, high","mae":0.560548657260194,"spearman":0.217102108111684,"r2_score":-0.30955373189927404,"validity":1,"rank":6}],"name":"Clearance Rat Liver Microsomes","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-clearance-human-hepatocytes":{"meta":{"title":"ISM Benchmarks: ADMET: Clearance Human Hepatocytes","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Intrinsic clearance in human hepatocytes captures both Phase I (CYP-mediated) and Phase II (conjugation) metabolism. Provides a more complete metabolic assessment than microsomes. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5641158123586562,"r2_score":-0.15773198681221645,"spearman":0.1508352941064932,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.7255695738348903,"r2_score":-0.5574994473468324,"spearman":0.0291739764979499,"validity":1,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.5712476334930577,"r2_score":-0.3221981854740954,"spearman":0.02164875938078792,"validity":0.9869281045751634,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":0.6065798799112435,"r2_score":-0.2067470442374557,"spearman":0.08278670712967635,"validity":1,"rank":3},{"model":"Kimi K3","mae":0.8379568377814376,"r2_score":-1.0684685199594646,"spearman":-0.06207615294446774,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.6279364843497907,"r2_score":-0.25227306376554903,"spearman":-0.015944379867164092,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":0.6935471291190644,"spearman":-0.13675659680100402,"r2_score":-0.587191364176884,"validity":0.9869281045751634,"rank":5},{"model":"GPT 5.6 Sol, high","mae":0.7110148246631249,"spearman":0.011769791989118043,"r2_score":-0.5841647380953727,"validity":1,"rank":6}],"name":"Clearance Human Hepatocytes","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-cyp1a2-inhibition-ic50":{"meta":{"title":"ISM Benchmarks: ADMET: CYP1A2 Inhibition IC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CYP1A2 inhibition potency (IC50) prediction. CYP1A2 metabolizes several important drugs including caffeine and theophylline; inhibition can cause clinically significant drug-drug interactions. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.4981182363031161,"r2_score":0.20828356574400508,"spearman":0.3408004658958973,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.6079025233380683,"r2_score":-0.17526740944527686,"spearman":0.2967053877015557,"validity":0.9971671388101983,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.45782917436519255,"r2_score":0.20028724901407846,"spearman":0.39475565452264333,"validity":0.9929178470254958,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.5232384422521247,"r2_score":-0.012262170491032842,"spearman":0.19570377772704886,"validity":1,"rank":3},{"model":"Kimi K3","mae":0.7287382874441041,"r2_score":-0.761733165183238,"spearman":0.09039686968347811,"validity":0.9249291784702549,"rank":8},{"model":"Grok 4.5, high","mae":0.5574695045750708,"r2_score":-0.10275454300409859,"spearman":0.11753057510717733,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.5418166130921985,"spearman":0.33532234758595225,"r2_score":-0.018031382338838053,"validity":0.9985835694050992,"rank":4},{"model":"GPT 5.6 Sol, high","mae":0.598950917254623,"spearman":0.30214843540799935,"r2_score":-0.18292317016778048,"validity":0.9957507082152974,"rank":6}],"name":"CYP1A2 Inhibition IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-cyp2c9-inhibition-ic50":{"meta":{"title":"ISM Benchmarks: ADMET: CYP2C9 Inhibition IC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CYP2C9 inhibition potency (IC50) prediction. CYP2C9 metabolizes ~15% of clinical drugs including warfarin and NSAIDs; inhibition poses significant safety risks. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.47087775383078734,"r2_score":0.004642634797133782,"spearman":0.2605796465764465,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.593420313172738,"r2_score":-0.4530144679943626,"spearman":0.2375532956977279,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.43915407761848346,"r2_score":0.037036134490034844,"spearman":0.3460490063593065,"validity":0.9917743830787309,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.5184790211202831,"r2_score":-0.19232808104626065,"spearman":0.2233843243258147,"validity":0.9964747356051704,"rank":3},{"model":"Kimi K3","mae":0.6666227399882075,"r2_score":-0.9624230161656675,"spearman":-0.006320601617490111,"validity":0.9964747356051704,"rank":8},{"model":"Grok 4.5, high","mae":0.5198775846180964,"r2_score":-0.17684189596377498,"spearman":0.01630966287926583,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":0.544230840787309,"spearman":0.2389301900091802,"r2_score":-0.3302501734320835,"validity":1,"rank":5},{"model":"GPT 5.6 Sol, high","mae":0.5952172224852769,"spearman":0.2574089124282712,"r2_score":-0.47132460099086515,"validity":0.9976498237367802,"rank":7}],"name":"CYP2C9 Inhibition IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-cyp2c19-inhibition-ic50":{"meta":{"title":"ISM Benchmarks: ADMET: CYP2C19 Inhibition IC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CYP2C19 inhibition potency (IC50) prediction. CYP2C19 is a highly polymorphic enzyme metabolizing proton pump inhibitors and clopidogrel. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.43847759020979027,"r2_score":0.030536116015021042,"spearman":0.2959509378876612,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.5527472452214452,"r2_score":-0.4100164922396712,"spearman":0.22991990455075878,"validity":1,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.4215933748815166,"r2_score":-0.0002524084163855189,"spearman":0.2882864788650176,"validity":0.9836829836829837,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.4750601644028104,"r2_score":-0.15870863986074868,"spearman":0.21008384677038697,"validity":0.9953379953379954,"rank":3},{"model":"Kimi K3","mae":0.6033730307692308,"r2_score":-0.8245275476570171,"spearman":0.10125906492721637,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.5025176414918414,"r2_score":-0.2082923137664925,"spearman":0.03736365623246327,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":0.5196772405594405,"spearman":0.25222411189568833,"r2_score":-0.35450919739109477,"validity":1,"rank":5},{"model":"GPT 5.6 Sol, high","mae":0.5269937533799535,"spearman":0.32951953506194437,"r2_score":-0.27020091643147626,"validity":1,"rank":6}],"name":"CYP2C19 Inhibition IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-cyp2d6-inhibition-ic50":{"meta":{"title":"ISM Benchmarks: ADMET: CYP2D6 Inhibition IC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CYP2D6 inhibition potency (IC50) prediction. CYP2D6 metabolizes ~25% of marketed drugs and is highly polymorphic in the population. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5103466097931034,"r2_score":0.1302817267223242,"spearman":0.44857626484764784,"validity":1,"rank":3},{"model":"GPT 5.6-sol","mae":0.6516805575966851,"r2_score":-0.4570964974299563,"spearman":0.3761616881479203,"validity":0.9986206896551724,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.4716357156944444,"r2_score":0.1897766735977683,"spearman":0.46445763450342753,"validity":0.993103448275862,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.5547896575966851,"r2_score":-0.07119709969009858,"spearman":0.34327833525297,"validity":0.9986206896551724,"rank":4},{"model":"Kimi K3","mae":0.6753501551724137,"r2_score":-0.49738450522730115,"spearman":0.22090829298426676,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.603415743862069,"r2_score":-0.09718690041686662,"spearman":0.15069151169445724,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.4998976029005526,"spearman":0.4673654950342024,"r2_score":0.11017708384807334,"validity":0.9986206896551724,"rank":2},{"model":"GPT 5.6 Sol, high","mae":0.6640732095303867,"spearman":0.3972838771270586,"r2_score":-0.500872309052327,"validity":0.9986206896551724,"rank":7}],"name":"CYP2D6 Inhibition IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-cyp3a4-inhibition-ic50":{"meta":{"title":"ISM Benchmarks: ADMET: CYP3A4 Inhibition IC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CYP3A4 inhibition potency (IC50) prediction. CYP3A4 is the most abundant hepatic CYP, metabolizing ~50% of marketed drugs. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6542249538391376,"r2_score":0.1449770348037034,"spearman":0.37408433768261606,"validity":1,"rank":3},{"model":"GPT 5.6-sol","mae":0.7660528688215767,"r2_score":-0.11351906211354579,"spearman":0.34388978128058884,"validity":0.9991708126036484,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.6533316688842632,"r2_score":-0.00018904911995054619,"spearman":0.36156123433719484,"validity":0.9958540630182421,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":0.6689585612697095,"r2_score":0.07221074113439263,"spearman":0.3579065856091741,"validity":0.9991708126036484,"rank":4},{"model":"Kimi K3","mae":0.9276437455472636,"r2_score":-0.655876431067788,"spearman":0.09977959317830547,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.7697206112189056,"r2_score":-0.09164339614303096,"spearman":0.05171811706225976,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.6509509582821577,"spearman":0.41968201484778334,"r2_score":0.0501680612425226,"validity":0.9991708126036484,"rank":1},{"model":"GPT 5.6 Sol, high","mae":0.7436276626285241,"spearman":0.3560153163666208,"r2_score":-0.041729415892914234,"validity":1,"rank":5}],"name":"CYP3A4 Inhibition IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-bcrp-inhibition":{"meta":{"title":"ISM Benchmarks: ADMET: BCRP Inhibition","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"Breast Cancer Resistance Protein (BCRP/ABCG2) inhibition prediction. BCRP is an efflux transporter affecting oral absorption, tissue distribution, and renal/biliary excretion of drugs. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6634707117282609,"r2_score":0.038757280270269256,"spearman":0.42606709949451954,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.6810262120036231,"r2_score":-0.04970146124237118,"spearman":0.3679716518750114,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mae":0.6460430422426471,"r2_score":0.10925614958408458,"spearman":0.4627801889143105,"validity":0.9855072463768116,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.7358759128079712,"r2_score":-0.23492679338395095,"spearman":0.39287630686156333,"validity":1,"rank":6},{"model":"Kimi K3","mae":0.797547116496377,"r2_score":-0.5797159020484428,"spearman":0.1625537120022468,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.7096187812862319,"r2_score":-0.08006134397943265,"spearman":0.16337378996437113,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.7868870931485508,"spearman":0.40012614731760476,"r2_score":-1.19880442773111,"validity":1,"rank":7},{"model":"GPT 5.6 Sol, high","mae":0.7005620970615943,"spearman":0.38213440230190593,"r2_score":-0.13595208682858262,"validity":1,"rank":4}],"name":"BCRP Inhibition","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-p-gp-inhibition":{"meta":{"title":"ISM Benchmarks: ADMET: P-gp Inhibition","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"P-glycoprotein (P-gp/MDR1) inhibition prediction. P-gp is a major efflux transporter; its inhibition can alter pharmacokinetics and cause drug-drug interactions affecting absorption and brain exposure. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6451608399698275,"r2_score":-0.020091610772812185,"spearman":0.38939528669710916,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.6835582676939655,"r2_score":-0.18779158390742334,"spearman":0.37439424591486975,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":0.5990938396869565,"r2_score":0.11065935998542842,"spearman":0.4682984019996356,"validity":0.9913793103448276,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.693642506538793,"r2_score":-0.24263291433213308,"spearman":0.4333971574129275,"validity":1,"rank":7},{"model":"Kimi K3","mae":0.7901167272456897,"r2_score":-0.5268528642722159,"spearman":0.2315792736114118,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.6761448355818966,"r2_score":-0.11057537478360246,"spearman":0.08349314749035167,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.6673816609698494,"spearman":0.46129545092262597,"r2_score":-0.23653792922385808,"validity":0.8577586206896551,"rank":4},{"model":"GPT 5.6 Sol, high","mae":0.6626823426853449,"spearman":0.42015599374322476,"r2_score":-0.09695998758142776,"validity":1,"rank":3}],"name":"P-gp Inhibition","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-hepg2-cc50":{"meta":{"title":"ISM Benchmarks: ADMET: HepG2 CC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CC50 (cytotoxic concentration 50%) in HepG2 hepatocellular carcinoma cells. A key indicator of hepatotoxicity risk and general cytotoxicity. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.3953354551091068,"r2_score":0.15176274570936865,"spearman":0.3367330358613752,"validity":1,"rank":2},{"model":"GPT 5.6-sol","mae":0.45152719512410733,"r2_score":-0.07437869837908506,"spearman":0.22414396591569788,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":0.3881478980227114,"r2_score":0.117729360131135,"spearman":0.36222142675048413,"validity":0.9954648526077098,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.434767399277798,"r2_score":-0.0462195164160506,"spearman":0.22595066926216326,"validity":0.9863945578231292,"rank":3},{"model":"Kimi K3","mae":0.5310967760269034,"r2_score":-0.4630586101847207,"spearman":0.06331805225497235,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.47995369356671935,"r2_score":-0.22492862382299283,"spearman":0.015159620712643256,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.43832430786887316,"spearman":0.2279551776944955,"r2_score":-0.07940205338047535,"validity":0.9977324263038548,"rank":4},{"model":"GPT 5.6 Sol, high","mae":0.45326317188489584,"spearman":0.23668047706503295,"r2_score":-0.09899504675478665,"validity":1,"rank":6}],"name":"HepG2 CC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-hek293-cc50":{"meta":{"title":"ISM Benchmarks: ADMET: HEK293 CC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"CC50 (cytotoxic concentration 50%) in HEK293 cells measures the concentration required to reduce cell viability by 50%. Important for assessing compound toxicity and therapeutic window. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6310594984101352,"r2_score":0.08869231423496404,"spearman":0.29495072771944936,"validity":1,"rank":1},{"model":"GPT 5.6-sol","mae":0.6938056089673527,"r2_score":-0.038809316958580586,"spearman":0.21814884004652477,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":0.6482335916360933,"r2_score":-0.05166324280390322,"spearman":0.24742937498048273,"validity":0.9828277046365197,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.6408492605504489,"r2_score":0.0805433454663389,"spearman":0.2718881965938064,"validity":0.9977103606182026,"rank":2},{"model":"Kimi K3","mae":0.8326239877747004,"r2_score":-0.536766602040605,"spearman":0.04558922196661535,"validity":0.9942759015455066,"rank":8},{"model":"Grok 4.5, high","mae":0.715070040576742,"r2_score":-0.10523820264279848,"spearman":0.06573774786783626,"validity":1,"rank":7},{"model":"Claude Opus 5, high","mae":0.6413426522612705,"spearman":0.25610527613689654,"r2_score":0.021900029309220526,"validity":0.9954207212364052,"rank":3},{"model":"GPT 5.6 Sol, high","mae":0.7022957443887645,"spearman":0.234782368878436,"r2_score":-0.050849555755500875,"validity":1,"rank":6}],"name":"HEK293 CC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-herg-ic50":{"meta":{"title":"ISM Benchmarks: ADMET: hERG IC50","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"hERG potassium channel inhibition potency (IC50) prediction. hERG inhibition causes QT prolongation and potentially fatal cardiac arrhythmias, making it a critical safety endpoint. Proprietary Chemistry42 data.","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6405610763004687,"r2_score":0.0014586481348217717,"spearman":0.40176316523284744,"validity":1,"rank":3},{"model":"GPT 5.6-sol","mae":0.7556515571146034,"r2_score":-0.2807152327747311,"spearman":0.3658005086269711,"validity":1,"rank":7},{"model":"Claude Opus 4.8, high","mae":0.5910640916291432,"r2_score":0.02360847836105484,"spearman":0.38922196206844695,"validity":0.9866666666666667,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":0.6546159347085513,"r2_score":-0.10679161118431524,"spearman":0.33505636027471347,"validity":0.9938461538461538,"rank":4},{"model":"Kimi K3","mae":0.914009690562981,"r2_score":-0.8503669241926497,"spearman":0.03232744295369731,"validity":1,"rank":8},{"model":"Grok 4.5, high","mae":0.6946206122893697,"r2_score":-0.1054525876128416,"spearman":0.035677218040588814,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.6336415488459388,"spearman":0.39095957427918365,"r2_score":-0.02460339969374581,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":0.7451969446132927,"spearman":0.38444839828930205,"r2_score":-0.27459661301738225,"validity":1,"rank":6}],"name":"hERG IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-aggregation-potential":{"meta":{"title":"ISM Benchmarks: ADMET: Aggregation potential","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"aggreg_bin_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","roc_auc_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.5758684903358376,"roc_auc_score":null,"validity":1,"rank":1},{"model":"GPT 5.6-sol","balanced_accuracy_score":0.5623434790984396,"roc_auc_score":null,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.5067431920474048,"roc_auc_score":null,"validity":0.9957507082152974,"rank":8},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.5192239773967765,"roc_auc_score":null,"validity":1,"rank":7},{"model":"Kimi K3","balanced_accuracy_score":0.5441050987715937,"roc_auc_score":null,"validity":0.9985835694050992,"rank":6},{"model":"Grok 4.5, high","balanced_accuracy_score":0.558314069222372,"roc_auc_score":null,"validity":1,"rank":3},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.5579771733851384,"roc_auc_score":null,"validity":0.9957507082152974,"rank":4},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.5563154177101394,"roc_auc_score":null,"validity":1,"rank":5}],"name":"Aggregation potential","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"roc_auc_score":{"label":"ROC AUC","shortLabel":"AUC","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-blood-brain-barrier-penetration":{"meta":{"title":"ISM Benchmarks: ADMET: Blood-Brain Barrier Penetration","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"bbb_bin_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","roc_auc_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7960507354067029,"roc_auc_score":null,"validity":1,"rank":2},{"model":"GPT 5.6-sol","balanced_accuracy_score":0.7775732479486972,"roc_auc_score":null,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.781797182916586,"roc_auc_score":null,"validity":0.9886444286728177,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.8097264629324399,"roc_auc_score":null,"validity":1,"rank":1},{"model":"Kimi K3","balanced_accuracy_score":0.7508612769866789,"roc_auc_score":null,"validity":1,"rank":7},{"model":"Grok 4.5, high","balanced_accuracy_score":0.7436389346735479,"roc_auc_score":null,"validity":1,"rank":8},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7774135970177545,"roc_auc_score":null,"validity":0.9985805535841022,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.7702896948252985,"roc_auc_score":null,"validity":0.9985805535841022,"rank":6}],"name":"Blood-Brain Barrier Penetration","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"roc_auc_score":{"label":"ROC AUC","shortLabel":"AUC","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-solubility-in-dmso":{"meta":{"title":"ISM Benchmarks: ADMET: Solubility in DMSO","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"dmso_sol_bin_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","roc_auc_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.5998006566604128,"roc_auc_score":null,"validity":1,"rank":1},{"model":"GPT 5.6-sol","balanced_accuracy_score":0.5651149155722326,"roc_auc_score":null,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.5363701775975576,"roc_auc_score":null,"validity":0.9981701738334858,"rank":7},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.5463632404181185,"roc_auc_score":null,"validity":1,"rank":6},{"model":"Kimi K3","balanced_accuracy_score":0.5334243500402037,"roc_auc_score":null,"validity":1,"rank":8},{"model":"Grok 4.5, high","balanced_accuracy_score":0.5575767220584293,"roc_auc_score":null,"validity":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.5937784776199411,"roc_auc_score":null,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.5466245644599304,"roc_auc_score":null,"validity":1,"rank":5}],"name":"Solubility in DMSO","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"roc_auc_score":{"label":"ROC AUC","shortLabel":"AUC","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-oatp1b1-inhibition":{"meta":{"title":"ISM Benchmarks: ADMET: OATP1B1 Inhibition","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"oatp1b1_bin_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","roc_auc_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7515720131749675,"roc_auc_score":null,"validity":1,"rank":1},{"model":"GPT 5.6-sol","balanced_accuracy_score":0.7039375187144425,"roc_auc_score":null,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.6998070478318269,"roc_auc_score":null,"validity":0.987460815047022,"rank":4},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.6859716538576704,"roc_auc_score":null,"validity":1,"rank":5},{"model":"Kimi K3","balanced_accuracy_score":0.6267454350161117,"roc_auc_score":null,"validity":0.4263322884012539,"rank":8},{"model":"Grok 4.5, high","balanced_accuracy_score":0.6690787503742888,"roc_auc_score":null,"validity":1,"rank":7},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7509481984229963,"roc_auc_score":null,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.685887402171169,"roc_auc_score":null,"validity":0.9968652037617555,"rank":6}],"name":"OATP1B1 Inhibition","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"roc_auc_score":{"label":"ROC AUC","shortLabel":"AUC","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-admet-ames":{"meta":{"title":"ISM Benchmarks: ADMET: Ames","category":"Medicinal Chemistry > ISM Benchmarks: ADMET","description":"ames_bin_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"admet_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","roc_auc_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7727883162665772,"roc_auc_score":null,"validity":1,"rank":2},{"model":"GPT 5.6-sol","balanced_accuracy_score":0.7647305690783952,"roc_auc_score":null,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7605021793996348,"roc_auc_score":null,"validity":0.9982788296041308,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.7942093810444875,"roc_auc_score":null,"validity":0.995697074010327,"rank":1},{"model":"Kimi K3","balanced_accuracy_score":0.731597280510324,"roc_auc_score":null,"validity":1,"rank":8},{"model":"Grok 4.5, high","balanced_accuracy_score":0.7504826254826255,"roc_auc_score":null,"validity":1,"rank":7},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7727287744632796,"roc_auc_score":null,"validity":0.9991394148020654,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.7631567903307034,"roc_auc_score":null,"validity":1,"rank":5}],"name":"Ames","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"roc_auc_score":{"label":"ROC AUC","shortLabel":"AUC","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-fgbench-functional-groups-reasoning-single-functional-group-direction":{"meta":{"title":"FGBench: Functional Groups Reasoning: Single Functional Group Direction","category":"Medicinal Chemistry > FGBench: Functional Groups Reasoning","description":"single_bool","metric_types":[],"paper_link":"#","source_file":"fgbench_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7723120168657766,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy_score":0.7765284609978917,"valid_fraction":1,"rank":2},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7645818692902319,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy_score":0.7730147575544624,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7652846099789178,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","accuracy_score":0.7806451612903226,"valid_fraction":0.9803232607167955,"rank":1}],"name":"Single Functional Group Direction","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-fgbench-functional-groups-reasoning-single-functional-group-value":{"meta":{"title":"FGBench: Functional Groups Reasoning: Single Functional Group Value","category":"Medicinal Chemistry > FGBench: Functional Groups Reasoning","description":"single_value","metric_types":[],"paper_link":"#","source_file":"fgbench_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["root_mean_squared_error","valid_fraction"],"leaderboard_metric":"root_mean_squared_error","results":[{"model":"GPT 5.5, medium","root_mean_squared_error":187.4382672946795,"valid_fraction":1,"rank":5},{"model":"Claude Opus 4.8, high","root_mean_squared_error":95.36496153820515,"valid_fraction":1,"rank":1},{"model":"Gemini 3.1 Pro, high","root_mean_squared_error":132.68714006657618,"valid_fraction":0.92,"rank":2},{"model":"Grok 4.5, high","root_mean_squared_error":178.24612154003552,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","root_mean_squared_error":188.632738094621,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","root_mean_squared_error":142.87379005655504,"valid_fraction":0.9946666666666667,"rank":3}],"name":"Single Functional Group Value","metric_info":{"root_mean_squared_error":{"label":"Root mean squared error","shortLabel":"RMSE","direction":"lower_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-fgbench-functional-groups-reasoning-functional-group-interaction-direction":{"meta":{"title":"FGBench: Functional Groups Reasoning: Functional Group Interaction Direction","category":"Medicinal Chemistry > FGBench: Functional Groups Reasoning","description":"interaction_bool","metric_types":[],"paper_link":"#","source_file":"fgbench_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7768421052631579,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy_score":0.7578947368421053,"valid_fraction":1,"rank":4},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7666900913562895,"valid_fraction":0.9985964912280701,"rank":3},{"model":"Grok 4.5, high","accuracy_score":0.7417543859649123,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7571929824561403,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","accuracy_score":0.7861230329041488,"valid_fraction":0.9810526315789474,"rank":1}],"name":"Functional Group Interaction Direction","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-fgbench-functional-groups-reasoning-functional-group-interaction-value":{"meta":{"title":"FGBench: Functional Groups Reasoning: Functional Group Interaction Value","category":"Medicinal Chemistry > FGBench: Functional Groups Reasoning","description":"interaction_value","metric_types":[],"paper_link":"#","source_file":"fgbench_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["root_mean_squared_error","valid_fraction"],"leaderboard_metric":"root_mean_squared_error","results":[{"model":"GPT 5.5, medium","root_mean_squared_error":56.0960070841589,"valid_fraction":1,"rank":5},{"model":"Claude Opus 4.8, high","root_mean_squared_error":52.979794082211086,"valid_fraction":1,"rank":3},{"model":"Gemini 3.1 Pro, high","root_mean_squared_error":51.53996174717839,"valid_fraction":0.9371428571428572,"rank":2},{"model":"Grok 4.5, high","root_mean_squared_error":48.5058083205647,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","root_mean_squared_error":64.19892799383588,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","root_mean_squared_error":53.37460039891855,"valid_fraction":0.9971428571428571,"rank":4}],"name":"Functional Group Interaction Value","metric_info":{"root_mean_squared_error":{"label":"Root mean squared error","shortLabel":"RMSE","direction":"lower_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-fgbench-functional-groups-reasoning-functional-group-comparison-direction":{"meta":{"title":"FGBench: Functional Groups Reasoning: Functional Group Comparison Direction","category":"Medicinal Chemistry > FGBench: Functional Groups Reasoning","description":"comparison_bool","metric_types":[],"paper_link":"#","source_file":"fgbench_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7896769662921348,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","accuracy_score":0.7801966292134831,"valid_fraction":1,"rank":3},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7622058306989814,"valid_fraction":0.9996488764044944,"rank":6},{"model":"Grok 4.5, high","accuracy_score":0.773876404494382,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7798455056179775,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy_score":0.7892882311486963,"valid_fraction":0.9964887640449438,"rank":2}],"name":"Functional Group Comparison Direction","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-fgbench-functional-groups-reasoning-functional-group-comparison-value":{"meta":{"title":"FGBench: Functional Groups Reasoning: Functional Group Comparison Value","category":"Medicinal Chemistry > FGBench: Functional Groups Reasoning","description":"comparison_value","metric_types":[],"paper_link":"#","source_file":"fgbench_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["root_mean_squared_error","valid_fraction"],"leaderboard_metric":"root_mean_squared_error","results":[{"model":"GPT 5.5, medium","root_mean_squared_error":77.75042463462997,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","root_mean_squared_error":71.44696022863349,"valid_fraction":1,"rank":1},{"model":"Gemini 3.1 Pro, high","root_mean_squared_error":80.30594784382524,"valid_fraction":0.9131034482758621,"rank":5},{"model":"Grok 4.5, high","root_mean_squared_error":80.99893316315011,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","root_mean_squared_error":77.25005157227898,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","root_mean_squared_error":76.35482398714886,"valid_fraction":0.9986206896551724,"rank":2}],"name":"Functional Group Comparison Value","metric_info":{"root_mean_squared_error":{"label":"Root mean squared error","shortLabel":"RMSE","direction":"lower_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht1a-ic50":{"meta":{"title":"GPCR Affinity Suite: 5-HT1A IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT1A IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9288298448775831,"r2_score":0.2896048451377756,"spearman":0.6163759658369017,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.3927608349851912,"r2_score":-0.8713465182610529,"spearman":0.17287124793223838,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":66.31060866607312,"r2_score":-343945.7482888639,"spearman":0.52840948585041,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.2045684444312166,"r2_score":-0.14951479666915235,"spearman":0.14145961101576707,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":22.75896566843828,"spearman":0.4037598605148591,"r2_score":-114747.97444284485,"validity":0.9978401727861771,"rank":5},{"model":"Claude Opus 5, high","mae":1.1972094415501224,"spearman":0.40571007040934165,"r2_score":-0.5837075441221451,"validity":1,"rank":2}],"name":"5-HT1A IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht1a-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT1A KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT1A KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9681548410811158,"r2_score":0.06623078294976414,"spearman":0.4417526697714633,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.30890449223486,"r2_score":-1.2891419703263223,"spearman":0.07818864296152443,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":94.42880375318352,"r2_score":-565483.9365080586,"spearman":0.3990029082850303,"validity":0.9994288977727013,"rank":6},{"model":"Grok 4.5, high","mae":1.1382959854307815,"r2_score":-0.22432277239827658,"spearman":0.05853407314367642,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":32.56048840148029,"spearman":0.3446262690263514,"r2_score":-142115.62848438838,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":4.878163222667058,"spearman":0.3571752505197261,"r2_score":-4618.540300053719,"validity":0.9988577955454027,"rank":4}],"name":"5-HT1A KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht1c-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT1C KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT1C KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9271871931291045,"r2_score":0.007962893705550256,"spearman":0.15433875269746858,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2879974637905647,"r2_score":-0.9393560723865297,"spearman":0.0689525214642585,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":27.393962529613038,"r2_score":-178086.05724947093,"spearman":0.07490627385686081,"validity":0.9993618379068283,"rank":5},{"model":"Grok 4.5, high","mae":1.0318840719664568,"r2_score":-0.1734746081249372,"spearman":0.08562182736056746,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":33.32017207520762,"spearman":0.06661817754664937,"r2_score":-222145.07682803134,"validity":0.9993618379068283,"rank":6},{"model":"Claude Opus 5, high","mae":1.2338987802753487,"spearman":0.07748350639796459,"r2_score":-0.9640664183380401,"validity":0.9993618379068283,"rank":3}],"name":"5-HT1C KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht2a-ic50":{"meta":{"title":"GPCR Affinity Suite: 5-HT2A IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT2A IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8927458530852821,"r2_score":0.07386805938159768,"spearman":0.39688011425121544,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.208572095002321,"r2_score":-1.0970211657982603,"spearman":0.19447238132221728,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":112.14296835624664,"r2_score":-839948.0746269233,"spearman":0.1280016132093862,"validity":0.9990215264187867,"rank":6},{"model":"Grok 4.5, high","mae":1.01571968278336,"r2_score":-0.2327318823784683,"spearman":0.12312769936283252,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":40.53624116314274,"spearman":0.08412440387016307,"r2_score":-303955.4788587256,"validity":0.9990215264187867,"rank":5},{"model":"Claude Opus 5, high","mae":4.62972504355425,"spearman":0.12149526348007167,"r2_score":-4226.276210639028,"validity":0.9990215264187867,"rank":4}],"name":"5-HT2A IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht2a-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT2A KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT2A KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9600812111953738,"r2_score":-0.1424693215080115,"spearman":0.3273246019924982,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.470669359464453,"r2_score":-2.3300277021750726,"spearman":0.0854149605125746,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":152.07371570146992,"r2_score":-962975.9528780719,"spearman":0.24650837062213796,"validity":0.999581589958159,"rank":6},{"model":"Grok 4.5, high","mae":1.0933807434177647,"r2_score":-0.22945757015696433,"spearman":0.06528925259653828,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":107.85347643772428,"spearman":0.1897238267427329,"r2_score":-622044.9998041668,"validity":0.999581589958159,"rank":5},{"model":"Claude Opus 5, high","mae":7.019575409432392,"spearman":0.2223632845783823,"r2_score":-31839.1052297569,"validity":0.9154811715481171,"rank":4}],"name":"5-HT2A KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht2b-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT2B KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT2B KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8198949928388266,"r2_score":0.07571356073345625,"spearman":0.2955694555260352,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.754712481205114,"r2_score":-3.867638186056805,"spearman":-0.09994705299479494,"validity":0.9923076923076923,"rank":5},{"model":"Gemini 3.1 Pro, high","mae":3.337972842685809,"r2_score":-2377.957218263133,"spearman":0.3081152745718679,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.9323129926589905,"r2_score":-0.1582530017995354,"spearman":0.08567556928831678,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.2185323481020558,"spearman":0.25610891059462165,"r2_score":-6.470427810400714,"validity":1,"rank":3},{"model":"Claude Opus 5, high","mae":1.587201099719592,"spearman":0.10233720365245672,"r2_score":-5.946506954710586,"validity":0.9942307692307693,"rank":4}],"name":"5-HT2B KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht5a-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT5A KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT5A KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7438009240150847,"r2_score":0.1914865749249206,"spearman":0.5012714095757649,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.9774319990559281,"r2_score":-0.4844858519358528,"spearman":0.2407640093165091,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.8320477142027941,"r2_score":-0.04162233089867984,"spearman":0.4025029750629075,"validity":0.9981949458483754,"rank":2},{"model":"Grok 4.5, high","mae":0.9265832314946504,"r2_score":-0.16251590165603336,"spearman":0.16953431394622342,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":1.076376009141691,"spearman":0.2015474601520834,"r2_score":-3.4062336873113326,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.0147000231243044,"spearman":0.26239102663498937,"r2_score":-0.6593493671475215,"validity":0.9981949458483754,"rank":5}],"name":"5-HT5A KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht6-ic50":{"meta":{"title":"GPCR Affinity Suite: 5-HT6 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT6 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8850291999181438,"r2_score":0.17859894868360005,"spearman":0.3580352625422662,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2090002578720256,"r2_score":-0.7877006147565475,"spearman":0.2243724440961746,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":62.11201080231498,"r2_score":-393808.5445382111,"spearman":0.24641695132717384,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.0843165748323964,"r2_score":-0.2836902949065252,"spearman":0.07104409453501134,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":92.70062501249127,"spearman":0.21816712633056232,"r2_score":-590863.40748868,"validity":0.9969604863221885,"rank":6},{"model":"Claude Opus 5, high","mae":1.8877130264484236,"spearman":0.1563575066807215,"r2_score":-17.91209193925991,"validity":1,"rank":4}],"name":"5-HT6 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht6-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT6 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT6 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9629149277416784,"r2_score":-0.05179942643424629,"spearman":0.25392909920384144,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2019237246179721,"r2_score":-0.9096481427641796,"spearman":0.024070789520023988,"validity":0.9720873786407767,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":22.208678528057284,"r2_score":-158709.35554911476,"spearman":0.2297077142823378,"validity":0.9995954692556634,"rank":6},{"model":"Grok 4.5, high","mae":0.9816913569192312,"r2_score":-0.21228160748498026,"spearman":0.07087052801235595,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":21.34099028271927,"spearman":0.13254350483684807,"r2_score":-131568.2323733265,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.5984971852928012,"spearman":0.24974908209628685,"r2_score":-79.84229232824164,"validity":0.9324433656957929,"rank":4}],"name":"5-HT6 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-5-ht7-ki":{"meta":{"title":"GPCR Affinity Suite: 5-HT7 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR 5-HT7 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8889026629362448,"r2_score":0.024913727963462695,"spearman":0.3774473577455043,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.162838729240789,"r2_score":-0.9067954713177253,"spearman":0.052463192481386575,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":52.67128924571521,"r2_score":-408697.0169110127,"spearman":0.3256164431611034,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.9862932669243439,"r2_score":-0.18434208717578682,"spearman":0.04246048932772581,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":39.16397495305815,"spearman":0.2580331419900572,"r2_score":-256882.94578935325,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":2.2034052979018837,"spearman":0.19383238021938162,"r2_score":-813.4893378491412,"validity":1,"rank":4}],"name":"5-HT7 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a1-ic50":{"meta":{"title":"GPCR Affinity Suite: A1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7120995444913477,"r2_score":-0.03201717718446795,"spearman":0.19181569869498982,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1276745793123402,"r2_score":-1.5018832234287371,"spearman":0.03422275950115041,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":108.90816413076185,"r2_score":-982023.8244356606,"spearman":-0.0005082953771154117,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.876109398588236,"r2_score":-0.3851179010650492,"spearman":-0.015688519778990294,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.6991152151553057,"spearman":0.0852888036595928,"r2_score":-31.21655087262978,"validity":0.9951456310679612,"rank":5},{"model":"Claude Opus 5, high","mae":1.0128362059023144,"spearman":0.09381966281834123,"r2_score":-11.480453432505504,"validity":0.1262135922330097,"rank":3}],"name":"A1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a1-ki":{"meta":{"title":"GPCR Affinity Suite: A1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9041886812872063,"r2_score":0.06650271237298466,"spearman":0.27054589756599207,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2677668145151815,"r2_score":-0.9682477447613989,"spearman":0.05676959188606392,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":56.06304156263957,"r2_score":-375930.3613509863,"spearman":0.20459995671836576,"validity":0.9987397605545053,"rank":6},{"model":"Grok 4.5, high","mae":1.033747889479256,"r2_score":-0.24969554410766537,"spearman":0.028586190801245775,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":12.775194039081697,"spearman":0.15469153533368363,"r2_score":-63360.78795885763,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.275115269764106,"spearman":0.09935522097952867,"r2_score":-1.0033693242193018,"validity":0.7441713925645873,"rank":4}],"name":"A1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a2a-ic50":{"meta":{"title":"GPCR Affinity Suite: A2A IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A2A IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8237705797119858,"r2_score":0.10285980892828905,"spearman":0.40811455127854324,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.032582145555581,"r2_score":-0.35276019846405093,"spearman":0.16140423819905325,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":22.17591478298515,"r2_score":-148550.65948489253,"spearman":0.25823364519662206,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.0256085426953707,"r2_score":-0.27250504219639615,"spearman":0.029017211042814794,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":13.981973376824325,"spearman":0.18846011373687419,"r2_score":-76213.9623010955,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.0033222260095582,"spearman":0.24823480077350552,"r2_score":-0.7810355289799285,"validity":1,"rank":2}],"name":"A2A IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a2a-ki":{"meta":{"title":"GPCR Affinity Suite: A2A KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A2A KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.0611847202928095,"r2_score":0.04975763113602816,"spearman":0.42570915477670185,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.367740642859818,"r2_score":-0.9059580146349022,"spearman":0.16852826356203163,"validity":0.9285975833028195,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":84.30752316504133,"r2_score":-441411.5461531858,"spearman":0.3615017003640385,"validity":0.9985353350421091,"rank":5},{"model":"Grok 4.5, high","mae":1.2124542664191893,"r2_score":-0.2294146720239565,"spearman":0.08964940808586133,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":84.69491637913275,"spearman":0.2681829133516085,"r2_score":-2543575.323459908,"validity":0.9996338337605273,"rank":6},{"model":"Claude Opus 5, high","mae":2.2769541147898784,"spearman":0.2646703033882301,"r2_score":-520.1681885358461,"validity":0.9080922738923471,"rank":4}],"name":"A2A KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a2b-ic50":{"meta":{"title":"GPCR Affinity Suite: A2B IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A2B IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7891635434542053,"r2_score":0.0719250047704999,"spearman":0.38452823080244763,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.9874624569350133,"r2_score":-0.45696131973007437,"spearman":0.15646280935905157,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.9865964917602186,"r2_score":-0.3842491899140583,"spearman":0.24473123131287997,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":0.9204046344846624,"r2_score":-0.18113820844442463,"spearman":0.11671476133171037,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":16.15196231876015,"spearman":0.09239255926787225,"r2_score":-125392.07206350773,"validity":0.9985228951255539,"rank":6},{"model":"Claude Opus 5, high","mae":1.1152263097510264,"spearman":0.1998897263139477,"r2_score":-0.52495127016755,"validity":0.8345642540620384,"rank":5}],"name":"A2B IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a2b-ki":{"meta":{"title":"GPCR Affinity Suite: A2B KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A2B KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7410178158035545,"r2_score":0.2590627308831469,"spearman":0.5445624132095618,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.098287708442116,"r2_score":-0.7042268542738528,"spearman":0.1721266497572668,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":61.68102106991023,"r2_score":-451455.6364283578,"spearman":0.41611085372562007,"validity":0.9952774498229043,"rank":5},{"model":"Grok 4.5, high","mae":1.007945608365221,"r2_score":-0.24448938608030724,"spearman":0.13627925465001042,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":219.14845524554974,"spearman":0.28486184133080894,"r2_score":-9858361.96497353,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":7.5413707569040564,"spearman":0.27482935234429845,"r2_score":-6878.736125733501,"validity":0.8866587957497049,"rank":4}],"name":"A2B KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-a3-ki":{"meta":{"title":"GPCR Affinity Suite: A3 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR A3 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8549020540459906,"r2_score":0.23740959145895535,"spearman":0.5315837502925416,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.266773414688389,"r2_score":-0.685830212018196,"spearman":0.1181927646489297,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":239.76628139076956,"r2_score":-1518365.46281311,"spearman":0.39886694321377403,"validity":0.9924686192468619,"rank":6},{"model":"Grok 4.5, high","mae":1.1057811295754667,"r2_score":-0.2505306382650112,"spearman":0.06714612546245982,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":18.05341225220089,"spearman":0.35357644308858455,"r2_score":-45413.49889296017,"validity":0.999163179916318,"rank":5},{"model":"Claude Opus 5, high","mae":1.4267439285620778,"spearman":0.268260585408494,"r2_score":-7.748775554155138,"validity":0.8527196652719665,"rank":4}],"name":"A3 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-alpha1a-ki":{"meta":{"title":"GPCR Affinity Suite: ALPHA1A KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR ALPHA1A KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9518105546619932,"r2_score":0.09373921238255067,"spearman":0.44421202182697106,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.3611857518294752,"r2_score":-1.2159863410416958,"spearman":0.099993039548771,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":17.900728507587907,"r2_score":-109614.28910700955,"spearman":0.3018715500191722,"validity":0.9966555183946488,"rank":4},{"model":"Grok 4.5, high","mae":1.0989225659240078,"r2_score":-0.21476449292721078,"spearman":0.09283354350158923,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":88.48311780065556,"spearman":0.25638764556728616,"r2_score":-2717466.0911424784,"validity":0.9983277591973244,"rank":6},{"model":"Claude Opus 5, high","mae":18.121153171158394,"spearman":0.2625029484501407,"r2_score":-108471.12763567075,"validity":1,"rank":5}],"name":"ALPHA1A KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-b1-ic50":{"meta":{"title":"GPCR Affinity Suite: B1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR B1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8774725018003467,"r2_score":0.04995050099541776,"spearman":0.36309927722004914,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0931258422698222,"r2_score":-0.529405659616806,"spearman":0.20853171783252958,"validity":0.9819819819819819,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.2973533620791333,"r2_score":-0.9245470204483555,"spearman":0.04899985343283514,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.0293381083519524,"r2_score":-0.2583336892115662,"spearman":0.023294557610053737,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":61.25490496442492,"spearman":0.0734385821230445,"r2_score":-454518.33441756456,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.2508904145625488,"spearman":0.08563379102428394,"r2_score":-1.3564853369845986,"validity":0.9819819819819819,"rank":4}],"name":"B1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-b1-ki":{"meta":{"title":"GPCR Affinity Suite: B1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR B1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9995904965579407,"r2_score":0.10819064205018636,"spearman":0.6166293611333439,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.187428892597349,"r2_score":-0.5497087506191061,"spearman":0.2294786064373304,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.348425793703088,"r2_score":-0.6561616521079685,"spearman":0.4118659554112682,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.1156881448772498,"r2_score":-0.20646310027445303,"spearman":0.17483528101697351,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":1.4703175120005199,"spearman":0.20165384460898292,"r2_score":-1.5338321149143095,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.0894061324077038,"spearman":0.3394034976649795,"r2_score":-0.39053328756777583,"validity":0.9923076923076923,"rank":2}],"name":"B1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-casr-ic50":{"meta":{"title":"GPCR Affinity Suite: CASR IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CASR IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9889771159288784,"r2_score":0.12828009076836067,"spearman":0.5307472986684818,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1814533193994492,"r2_score":-0.36388432734216125,"spearman":0.34611136176087354,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.4002130119591316,"r2_score":-0.7381120883129728,"spearman":0.30845491878517456,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.1299448461848547,"r2_score":-0.11694771491638423,"spearman":0.27497490072215747,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.4546320755507198,"spearman":0.23629826468300016,"r2_score":-1.5755099007374103,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.345751557366669,"spearman":0.3520697147859328,"r2_score":-0.9921412180437346,"validity":1,"rank":4}],"name":"CASR IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-cb1-ic50":{"meta":{"title":"GPCR Affinity Suite: CB1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CB1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.0489667456539111,"r2_score":0.19436638996816724,"spearman":0.5745171661688214,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.4183936464614915,"r2_score":-0.6180661327327495,"spearman":0.2013156468459512,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":44.14472470144679,"r2_score":-182140.3374139944,"spearman":0.4472922171028294,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.203734244242865,"r2_score":-0.07204253119550352,"spearman":0.17093024873370338,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":35.452374293568035,"spearman":0.4186146916352552,"r2_score":-128492.47383538034,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":2.0280138000416597,"spearman":0.3671066993198734,"r2_score":-51.44316059522895,"validity":0.9978540772532188,"rank":4}],"name":"CB1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-cb1-ki":{"meta":{"title":"GPCR Affinity Suite: CB1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CB1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.1002655262185037,"r2_score":0.0167921693592068,"spearman":0.4943077686188943,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.585766236785389,"r2_score":-0.95423024561064,"spearman":0.10594921618898699,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":172.8323111522865,"r2_score":-720443.1378881326,"spearman":0.43055387152102026,"validity":0.998638529611981,"rank":6},{"model":"Grok 4.5, high","mae":1.2620295770078385,"r2_score":-0.15097224426580236,"spearman":0.13802452508133292,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":33.10599190785755,"spearman":0.2869104635947398,"r2_score":-82243.2474331911,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":5.621606317842902,"spearman":0.25808375528548366,"r2_score":-1510.401246096327,"validity":0.9870660313138189,"rank":4}],"name":"CB1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-cb2-ic50":{"meta":{"title":"GPCR Affinity Suite: CB2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CB2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8194742123918706,"r2_score":0.22181417918066781,"spearman":0.4864108246815898,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.4563043504434512,"r2_score":-2.266316906685627,"spearman":0.09124582052936207,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":99.51695427637887,"r2_score":-695392.7227989504,"spearman":0.41657887900744583,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.0580885941621099,"r2_score":-0.2898102337467263,"spearman":0.07970541234427601,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.139116551757756,"spearman":0.2641328182295251,"r2_score":-0.8368478765576288,"validity":1,"rank":3},{"model":"Claude Opus 5, high","mae":3.250244739758966,"spearman":0.22137117002672785,"r2_score":-1382.2121813846557,"validity":1,"rank":5}],"name":"CB2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-cb2-ki":{"meta":{"title":"GPCR Affinity Suite: CB2 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CB2 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.0284677949336645,"r2_score":0.003026991900184739,"spearman":0.4055301707695396,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.4067819445749254,"r2_score":-0.9878220681291523,"spearman":0.08585076199895095,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":91.29686240830345,"r2_score":-539371.1385459589,"spearman":0.3881169718216028,"validity":0.9993117687543015,"rank":6},{"model":"Grok 4.5, high","mae":1.1592133019182396,"r2_score":-0.18417960512975395,"spearman":0.08189091750236514,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":19.90064626864719,"spearman":0.2827148874043812,"r2_score":-66734.14173723917,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":2.0136843972394565,"spearman":0.3354841308300141,"r2_score":-415.27494236552144,"validity":0.9986235375086029,"rank":4}],"name":"CB2 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ccr1-ic50":{"meta":{"title":"GPCR Affinity Suite: CCR1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CCR1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.0295346270704615,"r2_score":-0.11364536308066975,"spearman":0.3929882742662487,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.301658717120961,"r2_score":-0.9849034540228416,"spearman":0.0776981124133424,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.4566009759339051,"r2_score":-1.1650675065228349,"spearman":0.12928304321094303,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.033819533094573,"r2_score":-0.1748635381264756,"spearman":0.12421686571342488,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":14.483253993270205,"spearman":0.06500423850622601,"r2_score":-87738.40996653693,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.297238158895008,"spearman":0.15726448313180952,"r2_score":-1.1868221524724585,"validity":1,"rank":3}],"name":"CCR1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ccr2-ic50":{"meta":{"title":"GPCR Affinity Suite: CCR2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CCR2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7944449079817185,"r2_score":-0.04125535512466416,"spearman":0.19765066171013532,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1903212566693842,"r2_score":-1.6049237971192074,"spearman":0.027463210135183658,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.334619769251102,"r2_score":-1.6840275429272116,"spearman":0.11257474406877277,"validity":0.9969151670951156,"rank":5},{"model":"Grok 4.5, high","mae":0.8415338913781288,"r2_score":-0.2465462951208286,"spearman":0.054617381677084145,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":6.523220633418606,"spearman":0.04752578476457565,"r2_score":-52406.711798661054,"validity":0.9994858611825193,"rank":6},{"model":"Claude Opus 5, high","mae":1.3264430648066527,"spearman":0.09146441120594001,"r2_score":-2.522918738010034,"validity":0.9953727506426735,"rank":4}],"name":"CCR2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ccr3-ic50":{"meta":{"title":"GPCR Affinity Suite: CCR3 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CCR3 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.45882627992173397,"r2_score":-0.033575528381251774,"spearman":0.04840833755763866,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.8468238471841152,"r2_score":-1.448963209049921,"spearman":0.0696149669744195,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":8.888653303249448,"r2_score":-102956.45171159969,"spearman":0.08978494037469266,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.610906742938336,"r2_score":-0.36751537817116864,"spearman":0.11984284115216659,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":8.774641147733869,"spearman":0.08537790205564609,"r2_score":-102957.22511256294,"validity":0.9992101105845181,"rank":5},{"model":"Claude Opus 5, high","mae":0.8454497695188983,"spearman":0.13655203462701357,"r2_score":-1.9743739666036872,"validity":0.9928909952606635,"rank":3}],"name":"CCR3 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ccr5-ic50":{"meta":{"title":"GPCR Affinity Suite: CCR5 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CCR5 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7087827769164743,"r2_score":0.015668698292941707,"spearman":0.2526944335122455,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1732055478142487,"r2_score":-1.445999861698009,"spearman":0.10642966890347114,"validity":0.9942028985507246,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.4403596237889462,"r2_score":-2.0944621776659407,"spearman":0.10478535992980467,"validity":1,"rank":4},{"model":"Grok 4.5, high","mae":0.9622740691322419,"r2_score":-0.5214949664546384,"spearman":0.0639616764165644,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":34.713694167177295,"spearman":0.04762328536807284,"r2_score":-248550.0341954506,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.4613313043977192,"spearman":0.05137385135189924,"r2_score":-2.9399793166384782,"validity":0.9782608695652174,"rank":5}],"name":"CCR5 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-crhr1-ic50":{"meta":{"title":"GPCR Affinity Suite: CRHR1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CRHR1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.544814222004591,"r2_score":0.12330239789229269,"spearman":0.480491878339313,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.8519832617985479,"r2_score":-1.1329154896284086,"spearman":0.19068206708584476,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.294967590090019,"r2_score":-2.953060351724486,"spearman":0.11524655372055416,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":0.7435863058301275,"r2_score":-0.3253121382523678,"spearman":0.15098080918742324,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":16.903352377992636,"spearman":0.17206400830354346,"r2_score":-181187.56878643698,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.2010694405562896,"spearman":0.14593851169367406,"r2_score":-10.790330781976026,"validity":0.8624708624708625,"rank":4}],"name":"CRHR1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-crhr1-ki":{"meta":{"title":"GPCR Affinity Suite: CRHR1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CRHR1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8419351626058166,"r2_score":-0.06678342547993066,"spearman":0.302280898546231,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0574821621748491,"r2_score":-0.7814784131238786,"spearman":0.1184361223242672,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":120.53976918034516,"r2_score":-1159356.850312217,"spearman":0.09914827227354045,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.9441852575412997,"r2_score":-0.35045991112132135,"spearman":0.09535574014247444,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":26.571449850626742,"spearman":0.1851818253525243,"r2_score":-206691.78526214554,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":6.840833874948398,"spearman":0.2154776717936252,"r2_score":-5475.174112061295,"validity":0.8840336134453781,"rank":4}],"name":"CRHR1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-cxcr3-ic50":{"meta":{"title":"GPCR Affinity Suite: CXCR3 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CXCR3 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7192396350002533,"r2_score":-0.057530760177997076,"spearman":0.38873571410630653,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.8228320938855692,"r2_score":-0.22103673201243956,"spearman":0.28672952704074095,"validity":0.9982547993019197,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":0.8268797747929452,"r2_score":-0.3366086522686811,"spearman":0.33757268832384224,"validity":0.9982547993019197,"rank":4},{"model":"Grok 4.5, high","mae":0.7840770593484537,"r2_score":-0.04137178245417239,"spearman":0.20747578226703153,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.0189721233642317,"spearman":0.24843753694181175,"r2_score":-0.8333589880596881,"validity":0.9965095986038395,"rank":6},{"model":"Claude Opus 5, high","mae":0.8621058043160678,"spearman":0.29900101615241453,"r2_score":-0.511604915347261,"validity":0.9790575916230366,"rank":5}],"name":"CXCR3 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-cxcr4-ic50":{"meta":{"title":"GPCR Affinity Suite: CXCR4 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR CXCR4 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9088101979838162,"r2_score":0.3064828911685068,"spearman":0.6307063123246902,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.4649400339234508,"r2_score":-1.104743766725771,"spearman":0.22132516050383952,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.3655477513776881,"r2_score":-0.7504497973694788,"spearman":0.40376042866369105,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":1.192206138692996,"r2_score":-0.11111010471410943,"spearman":0.2555868016213193,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":47.35078566940644,"spearman":0.3809933214243473,"r2_score":-187707.5256324776,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.6867878007358885,"spearman":0.2206492736212647,"r2_score":-1.6854096857727154,"validity":0.9824561403508771,"rank":5}],"name":"CXCR4 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-d1-ki":{"meta":{"title":"GPCR Affinity Suite: D1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR D1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7584374587931821,"r2_score":0.07369813049996743,"spearman":0.4581234794808841,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1040744394020559,"r2_score":-0.9364264905343358,"spearman":0.07807397264029245,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.978149119046407,"r2_score":-0.6559918188695375,"spearman":0.41831157442514677,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":0.9138813055171574,"r2_score":-0.23727262551281192,"spearman":0.06396054668225044,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":70.22926343822556,"spearman":0.35657678110867713,"r2_score":-650044.3933776766,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.1862482473251148,"spearman":0.22972537108398242,"r2_score":-2.332139934424375,"validity":1,"rank":5}],"name":"D1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-d2-ic50":{"meta":{"title":"GPCR Affinity Suite: D2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR D2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7160490856022345,"r2_score":0.27904570121908023,"spearman":0.6268566155073415,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0700189229971055,"r2_score":-0.8076827143426843,"spearman":0.27434138732403973,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":61.85194742511711,"r2_score":-497660.90336180956,"spearman":0.30489487286821754,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.9402946802618727,"r2_score":-0.15065866535591121,"spearman":0.23694598010406245,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":21.676229793223026,"spearman":0.2840768774446072,"r2_score":-166086.53253079508,"validity":0.9979959919839679,"rank":4},{"model":"Claude Opus 5, high","mae":21.835989810373466,"spearman":0.26828130721044663,"r2_score":-166053.16777500225,"validity":1,"rank":5}],"name":"D2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-d2-ki":{"meta":{"title":"GPCR Affinity Suite: D2 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR D2 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7377721343719934,"r2_score":0.04869561398977862,"spearman":0.30344714632372777,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.093168832436392,"r2_score":-1.3256312264925065,"spearman":0.07648943300493195,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":83.44400642546955,"r2_score":-846972.6548511046,"spearman":0.24037080419417545,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.8461745488479903,"r2_score":-0.23510651162801288,"spearman":0.05353418005872363,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":28.33656860610849,"spearman":0.18994172508553947,"r2_score":-224729.7561393602,"validity":0.9995751911639762,"rank":5},{"model":"Claude Opus 5, high","mae":1.2326106618376316,"spearman":0.21263229004449888,"r2_score":-41.23917553779363,"validity":1,"rank":4}],"name":"D2 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-d3-ki":{"meta":{"title":"GPCR Affinity Suite: D3 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR D3 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.754180194710641,"r2_score":0.02529732103767912,"spearman":0.2804616205983196,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":2.7829104553462853,"r2_score":-8014.849608634087,"spearman":0.056480055306927866,"validity":0.976161738080869,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":57.636146265309584,"r2_score":-563773.9530357305,"spearman":0.2509230479643419,"validity":0.9993964996982498,"rank":6},{"model":"Grok 4.5, high","mae":0.8497225960299725,"r2_score":-0.3152396401296904,"spearman":0.0584366696856545,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":12.408088807529456,"spearman":0.1791913663271693,"r2_score":-94363.8983618525,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":2.946540963387565,"spearman":0.15944021192534769,"r2_score":-1496.9298914834537,"validity":0.995775497887749,"rank":4}],"name":"D3 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-d4-ki":{"meta":{"title":"GPCR Affinity Suite: D4 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR D4 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9820467557266056,"r2_score":-0.24711005183468004,"spearman":0.15385805505892,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","mae":1.1302138841693672,"r2_score":-0.7208008803286217,"spearman":0.027349894724625377,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":14.776844459638161,"r2_score":-114711.41007903941,"spearman":0.22810909511930322,"validity":0.9986357435197817,"rank":6},{"model":"Grok 4.5, high","mae":0.9798784410229366,"r2_score":-0.2254350491388799,"spearman":0.05977415448492073,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mae":4.6299347422892145,"spearman":0.17391061066004476,"r2_score":-4142.547425344634,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.1466813368940802,"spearman":0.0662137584443289,"r2_score":-1.0442928071557596,"validity":1,"rank":4}],"name":"D4 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ep2-ic50":{"meta":{"title":"GPCR Affinity Suite: EP2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR EP2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9080884830771124,"r2_score":-0.2295356158064903,"spearman":0.13223798264351422,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","mae":1.0484731757403103,"r2_score":-0.5575865683908954,"spearman":0.040221788392356905,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.047772607247186,"r2_score":-0.7542920682313268,"spearman":0.11000967036000825,"validity":0.988950276243094,"rank":3},{"model":"Grok 4.5, high","mae":0.899185329038736,"r2_score":-0.1390599509233208,"spearman":0.13813822975984547,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mae":1.2155960983101024,"spearman":0.08253224675608493,"r2_score":-2.579907914688779,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.4938769422329279,"spearman":0.10457385187829636,"r2_score":-125.68122274918031,"validity":1,"rank":6}],"name":"EP2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ep4-ic50":{"meta":{"title":"GPCR Affinity Suite: EP4 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR EP4 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7187178909074203,"r2_score":0.09037286692111246,"spearman":0.4057512278731049,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.8606303851986613,"r2_score":-0.34079440065951916,"spearman":0.16798303191741945,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":15.362127885400401,"r2_score":-151491.01461916656,"spearman":0.2915491333977615,"validity":0.9985632183908046,"rank":5},{"model":"Grok 4.5, high","mae":0.8381682893572054,"r2_score":-0.18718031944115499,"spearman":0.15787482690002216,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":29.83941541452227,"spearman":0.24523160329404534,"r2_score":-302718.72170711803,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":0.86074488349322,"spearman":0.26773829063401383,"r2_score":-0.4019776548873679,"validity":1,"rank":4}],"name":"EP4 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-eta-ic50":{"meta":{"title":"GPCR Affinity Suite: ETA IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR ETA IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.0678307220725038,"r2_score":-0.03232913416606209,"spearman":0.4419925974635248,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.3283020134649899,"r2_score":-0.42083727024054984,"spearman":0.25003006388939136,"validity":0.17379679144385027,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":4.4012731699759025,"r2_score":-1993.417450127942,"spearman":0.34314670695227273,"validity":1,"rank":4},{"model":"Grok 4.5, high","mae":1.1975950706702074,"r2_score":-0.1997489242672792,"spearman":0.15058574125341279,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":111.99552929419465,"spearman":0.32924749031672923,"r2_score":-2223918.150899553,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":6.554907731495437,"spearman":0.3105129333230441,"r2_score":-3318.4382694120272,"validity":0.6310160427807486,"rank":5}],"name":"ETA IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ghs-r-ic50":{"meta":{"title":"GPCR Affinity Suite: GHS-R IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GHS-R IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9527455247489632,"r2_score":-0.1288960095553433,"spearman":0.14512432146694468,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1886400549601264,"r2_score":-0.9627280260657194,"spearman":0.08923866284965001,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":16.883979056899054,"r2_score":-124544.37946299317,"spearman":0.20695595781617016,"validity":0.9952830188679245,"rank":6},{"model":"Grok 4.5, high","mae":1.0040657985244186,"r2_score":-0.23702154804529108,"spearman":0.05893393214859655,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.353463799988669,"spearman":0.07931116874997339,"r2_score":-2.576555432142009,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.3334747305540264,"spearman":-0.008347569366447663,"r2_score":-1.5002273882893835,"validity":0.9984276729559748,"rank":4}],"name":"GHS-R IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-glucagon-ic50":{"meta":{"title":"GPCR Affinity Suite: GLUCAGON IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GLUCAGON IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9816247116309689,"r2_score":-0.34487275281182384,"spearman":0.19474903129750756,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mae":1.0402661825227113,"r2_score":-0.8623940001381312,"spearman":0.1136318391640091,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.0563638041729404,"r2_score":-0.9290291296944686,"spearman":0.1995976806503804,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":0.8261360643011303,"r2_score":-0.3553102558674639,"spearman":0.11524408957557121,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mae":1.2602559741281898,"spearman":0.10780170184123274,"r2_score":-1.457933395289316,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":0.8696778394556226,"spearman":0.22689740921872018,"r2_score":-0.4228449841461652,"validity":0.9987129987129987,"rank":2}],"name":"GLUCAGON IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-glucagon-ki":{"meta":{"title":"GPCR Affinity Suite: GLUCAGON KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GLUCAGON KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5085476927088184,"r2_score":0.16383945556088386,"spearman":0.5081070456032979,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.7376513861710412,"r2_score":-0.49010282361811397,"spearman":0.24643545284360244,"validity":1,"rank":5},{"model":"Gemini 3.1 Pro, high","mae":0.6377924600711538,"r2_score":-0.029313292875321695,"spearman":0.3640654278581826,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":0.5881354831876938,"r2_score":-0.03731621648658723,"spearman":0.15710749878298863,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":0.8400282762828574,"spearman":0.2996689240325983,"r2_score":-0.8954740506711423,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":0.6631824119918532,"spearman":0.17245870451361753,"r2_score":-0.1312306741112832,"validity":1,"rank":4}],"name":"GLUCAGON KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-gnrhr-ic50":{"meta":{"title":"GPCR Affinity Suite: GNRHR IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GNRHR IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9355251932756086,"r2_score":0.27649083196602153,"spearman":0.6011943542936281,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.354872967521309,"r2_score":-0.578330508113047,"spearman":0.15827953285196633,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":4.000770057622623,"r2_score":-1305.900231130798,"spearman":0.24125667565321976,"validity":0.9975429975429976,"rank":5},{"model":"Grok 4.5, high","mae":1.194216874375882,"r2_score":-0.21661561090145964,"spearman":0.08368579409451321,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":4.60343607101733,"spearman":0.20591721913573072,"r2_score":-1091.7966263640733,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.3395476260404642,"spearman":0.21005786457324174,"r2_score":-0.631362778040333,"validity":0.9508599508599509,"rank":3}],"name":"GNRHR IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-gnrhr-ki":{"meta":{"title":"GPCR Affinity Suite: GNRHR KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GNRHR KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9227596898908738,"r2_score":-0.3182478960455397,"spearman":0.2713741293895393,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0712318287052998,"r2_score":-0.6163824986078026,"spearman":0.17794870010524122,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.4098366674947767,"r2_score":-17.62731803498159,"spearman":0.16105755532209606,"validity":1,"rank":4},{"model":"Grok 4.5, high","mae":0.9905043012434647,"r2_score":-0.3015863302626163,"spearman":0.11103056559137814,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":49.62492809113426,"spearman":0.16491990705366139,"r2_score":-364871.43546283024,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":11.849420157491643,"spearman":0.2620308608235254,"r2_score":-19492.62439564306,"validity":0.9566265060240964,"rank":5}],"name":"GNRHR KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-gpr44-ic50":{"meta":{"title":"GPCR Affinity Suite: GPR44 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GPR44 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8483903463670025,"r2_score":-0.10053327615234009,"spearman":0.2793546129050092,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.9985992996746639,"r2_score":-0.5530829638221888,"spearman":0.02871851284043663,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.9748669008418966,"r2_score":-938.8623891522547,"spearman":0.1254745927601845,"validity":0.9990808823529411,"rank":5},{"model":"Grok 4.5, high","mae":0.9204710577547649,"r2_score":-0.28562447107975575,"spearman":0.09785758551638872,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":3.9403623404428516,"spearman":0.11547862014257536,"r2_score":-8453.562621878958,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.043717513088715,"spearman":0.15944870427203106,"r2_score":-1.0988783595500449,"validity":1,"rank":4}],"name":"GPR44 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-gpr44-ki":{"meta":{"title":"GPCR Affinity Suite: GPR44 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR GPR44 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6263024106144314,"r2_score":0.22280565446378708,"spearman":0.5560410723518643,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.9197249832749173,"r2_score":-0.3740540469503364,"spearman":0.2617369059790111,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":2.4364093545854892,"r2_score":-1140.9982995721912,"spearman":0.25358562754092445,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":0.8471706108234233,"r2_score":-0.14746069074833423,"spearman":0.31897088963810566,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":51.67796708839232,"spearman":0.34025140963489536,"r2_score":-454294.7502953222,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.306614848404762,"spearman":0.25025427655225707,"r2_score":-70.68440780174028,"validity":1,"rank":4}],"name":"GPR44 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-il8rb-ic50":{"meta":{"title":"GPCR Affinity Suite: IL8RB IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR IL8RB IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8106448734924557,"r2_score":0.2678818602421249,"spearman":0.5498752832722609,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.319334325710672,"r2_score":-1.1182398824541253,"spearman":0.1541831667237666,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":23.408334962237287,"r2_score":-147734.28563014968,"spearman":0.31343994344245707,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.0562691359170595,"r2_score":-0.2213919436821874,"spearman":0.2296254347490622,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.4165354409461215,"spearman":0.19581824291943956,"r2_score":-1.7754646302254038,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":1.6237906799396427,"spearman":0.2400211809480327,"r2_score":-2.642118550927413,"validity":0.9824945295404814,"rank":5}],"name":"IL8RB IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-machrm1-ic50":{"meta":{"title":"GPCR Affinity Suite: MACHRM1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MACHRM1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9287898950683221,"r2_score":0.31318255657511174,"spearman":0.5735734163289824,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.5705247654514096,"r2_score":-0.9931455111605463,"spearman":0.10684703593295095,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":44.62057290766808,"r2_score":-170890.84459118254,"spearman":0.3479680524360305,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.2158363489344501,"r2_score":-0.20385235316980888,"spearman":0.14080399811120528,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":44.68137565957228,"spearman":0.3189752581866943,"r2_score":-170903.77821669795,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.8893295727253618,"spearman":0.11269816909488399,"r2_score":-2.177489752306503,"validity":1,"rank":4}],"name":"MACHRM1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-machrm1-ki":{"meta":{"title":"GPCR Affinity Suite: MACHRM1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MACHRM1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9464720070447298,"r2_score":0.29277348164412476,"spearman":0.6422010000098625,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":2.0500012096148783,"r2_score":-1.799575863105959,"spearman":0.07623132306577599,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":43.72625803504143,"r2_score":-120987.10075261613,"spearman":0.4331168747726971,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.372859280906149,"r2_score":-0.2346500533165119,"spearman":0.09707197207070016,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":4.439848807204448,"spearman":0.3666405612224296,"r2_score":-943.6247693009676,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":2.31402226329111,"spearman":0.15480210836710165,"r2_score":-2.4194695942509017,"validity":0.9936575052854123,"rank":4}],"name":"MACHRM1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-machrm2-ki":{"meta":{"title":"GPCR Affinity Suite: MACHRM2 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MACHRM2 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.031869759032369,"r2_score":0.14637167030717546,"spearman":0.44978435332323385,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.6764316524488798,"r2_score":-1.172344839287684,"spearman":0.14187444467338414,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":2.8779023147214433,"r2_score":-509.7217858127575,"spearman":0.40922785082090474,"validity":0.9985734664764622,"rank":4},{"model":"Grok 4.5, high","mae":1.2493365021501737,"r2_score":-0.12868562315912624,"spearman":0.07338061125563232,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":88.7849253668093,"spearman":0.33986102158703674,"r2_score":-329440.04684561875,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":3.85651510972878,"spearman":0.18366731853367893,"r2_score":-551.0924907713024,"validity":0.9971469329529244,"rank":5}],"name":"MACHRM2 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-machrm3-ki":{"meta":{"title":"GPCR Affinity Suite: MACHRM3 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MACHRM3 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.077272206470929,"r2_score":0.22278716045101932,"spearman":0.5523327219229879,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.7468255245696522,"r2_score":-1.6448097807911273,"spearman":0.13533059348606458,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":58.35062203162305,"r2_score":-220040.23130024693,"spearman":0.4481397443231332,"validity":0.9988814317673378,"rank":5},{"model":"Grok 4.5, high","mae":1.3303979274858426,"r2_score":-0.15404149878934548,"spearman":0.14311722393887888,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":60.33827117002672,"spearman":0.393470461140476,"r2_score":-220646.9716291194,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":4.064932234934847,"spearman":0.39465832738288664,"r2_score":-903.9083374412337,"validity":0.9910514541387024,"rank":4}],"name":"MACHRM3 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-machrm4-ic50":{"meta":{"title":"GPCR Affinity Suite: MACHRM4 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MACHRM4 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9848293924450608,"r2_score":0.3399705758480328,"spearman":0.6482926870365796,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2789492271075582,"r2_score":-0.4104753554430658,"spearman":0.3836687902847252,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.3028552809138532,"r2_score":-0.3965808555017625,"spearman":0.3805533463379561,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.2454690505791655,"r2_score":-0.14048284878635986,"spearman":0.29228913163678666,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":46.968867521609745,"spearman":0.3192128645243619,"r2_score":-210070.45137689816,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.2988322905321017,"spearman":0.42098323332060633,"r2_score":-0.3704899934504533,"validity":1,"rank":4}],"name":"MACHRM4 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-machrm4-ki":{"meta":{"title":"GPCR Affinity Suite: MACHRM4 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MACHRM4 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9415812601139639,"r2_score":0.2651359308705946,"spearman":0.6454407327231625,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.915523225765286,"r2_score":-1.3982653629174897,"spearman":0.09066180545547511,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":198.18904127208617,"r2_score":-605160.3979468336,"spearman":0.3949819775598095,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.5165672727016994,"r2_score":-0.29115149792683037,"spearman":0.026011054128468443,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":161.1937681586261,"spearman":0.317359859159147,"r2_score":-484270.52057640743,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":6.131304144222075,"spearman":0.18441614122756547,"r2_score":-1203.6982867370532,"validity":1,"rank":4}],"name":"MACHRM4 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mc3r-ki":{"meta":{"title":"GPCR Affinity Suite: MC3R KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MC3R KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6402799547003857,"r2_score":0.29883627671801594,"spearman":0.514285794190519,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0619165345413986,"r2_score":-0.8983984317388722,"spearman":0.2214055990406215,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":88.74504364033213,"r2_score":-814616.9903958371,"spearman":0.355513237002062,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.0037293314249915,"r2_score":-0.7831575081636724,"spearman":0.04561139376979801,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":126.11592680197406,"spearman":0.33394409760744515,"r2_score":-469586.9421594997,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":10.740216845633585,"spearman":0.3945518715949856,"r2_score":-10035.277777108744,"validity":0.875968992248062,"rank":4}],"name":"MC3R KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mc4r-ic50":{"meta":{"title":"GPCR Affinity Suite: MC4R IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MC4R IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7303589421167608,"r2_score":0.3485903075629666,"spearman":0.6116554785407674,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.9573078021908158,"r2_score":-0.06846246122029553,"spearman":0.3972495674150626,"validity":1,"rank":2},{"model":"Gemini 3.1 Pro, high","mae":1.042510326646723,"r2_score":-0.8240245302892051,"spearman":0.38814045760779486,"validity":0.994475138121547,"rank":4},{"model":"Grok 4.5, high","mae":1.0138912056906275,"r2_score":-0.14636052558092016,"spearman":0.17127379543249327,"validity":0.9972375690607734,"rank":3},{"model":"GPT 5.6 Sol, high","mae":1.6821363535780929,"spearman":0.19088014593814132,"r2_score":-57.442210247072495,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.1106593527951256,"spearman":0.34455208093239753,"r2_score":-0.8828971064070041,"validity":0.8480662983425414,"rank":5}],"name":"MC4R IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mc4r-ki":{"meta":{"title":"GPCR Affinity Suite: MC4R KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MC4R KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9787079867808296,"r2_score":-0.09448734866144659,"spearman":0.3251518186055468,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2290789487022427,"r2_score":-0.9024809967766363,"spearman":-0.014608050713927087,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.4869913782191106,"r2_score":-27.696303723435054,"spearman":0.10814589146521722,"validity":1,"rank":4},{"model":"Grok 4.5, high","mae":1.0837346148372626,"r2_score":-0.39267968195134295,"spearman":-0.0286246305181404,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.5843490600605354,"spearman":0.12821233393716142,"r2_score":-5.518371931613592,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":6.151331785614073,"spearman":0.16400229235275346,"r2_score":-3811.5211323939875,"validity":0.777511961722488,"rank":6}],"name":"MC4R KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mc5r-ki":{"meta":{"title":"GPCR Affinity Suite: MC5R KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MC5R KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6104281431296441,"r2_score":0.3559348420553241,"spearman":0.6854758980510623,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1500312011122156,"r2_score":-1.3293806319621373,"spearman":0.16439473745545857,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.7235734650109805,"r2_score":0.13198105052896247,"spearman":0.45874886344810556,"validity":1,"rank":2},{"model":"Grok 4.5, high","mae":0.994687765577542,"r2_score":-0.6252607353562893,"spearman":0.08859248094244518,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":60.74060112311158,"spearman":0.35311702528693556,"r2_score":-292771.4575257562,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":63.48249818303093,"spearman":0.40044097356047387,"r2_score":-333718.1571692895,"validity":0.9195402298850575,"rank":6}],"name":"MC5R KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mglur1-ic50":{"meta":{"title":"GPCR Affinity Suite: MGLUR1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MGLUR1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8632221402692267,"r2_score":0.15965727189972034,"spearman":0.5764590915325419,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.4745808499359079,"r2_score":-2.083243581531059,"spearman":0.26622878906634545,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":29.37242382905559,"r2_score":-120710.18028837412,"spearman":0.1983235026234127,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.1361963732460136,"r2_score":-0.19310927935374966,"spearman":0.2424126806002999,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.3838138684810055,"spearman":0.2388136940830043,"r2_score":-1.1201002812886864,"validity":0.9943977591036415,"rank":3},{"model":"Claude Opus 5, high","mae":1.4829450682024268,"spearman":0.26311899348685397,"r2_score":-1.1113668870952305,"validity":0.9859943977591037,"rank":5}],"name":"MGLUR1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mglur2-ic50":{"meta":{"title":"GPCR Affinity Suite: MGLUR2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MGLUR2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.860575507422542,"r2_score":-0.19961036420267364,"spearman":0.3612736590962495,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0859390875082569,"r2_score":-0.8289267340283297,"spearman":0.12688383420912225,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.2341274437551697,"r2_score":-15.249995991133858,"spearman":0.1077586889125334,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.9323309023779758,"r2_score":-0.23228927933417287,"spearman":0.11268523919168574,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.166966145130951,"spearman":0.05275954968337992,"r2_score":-1.1907171951364472,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.031720177714729,"spearman":0.1376835972321783,"r2_score":-0.6986374653539069,"validity":0.9403846153846154,"rank":3}],"name":"MGLUR2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mglur5-ic50":{"meta":{"title":"GPCR Affinity Suite: MGLUR5 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MGLUR5 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.060316896157771,"r2_score":-0.05520405036708609,"spearman":0.3817838025114994,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.644960113446269,"r2_score":-1.6755316194334622,"spearman":0.09296753780013087,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.4466806989552332,"r2_score":-1.0696143095512123,"spearman":0.22037950462551684,"validity":0.9986595174262735,"rank":3},{"model":"Grok 4.5, high","mae":1.191214121218785,"r2_score":-0.27899361406654166,"spearman":0.07383253788597034,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":12.460064049091937,"spearman":0.18605191593654083,"r2_score":-52919.42969163751,"validity":0.9986595174262735,"rank":6},{"model":"Claude Opus 5, high","mae":1.651172199814695,"spearman":0.14045252137952696,"r2_score":-4.366962246675535,"validity":0.9986595174262735,"rank":5}],"name":"MGLUR5 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-mglur5-ki":{"meta":{"title":"GPCR Affinity Suite: MGLUR5 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR MGLUR5 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.0656368609662847,"r2_score":0.10405860241466203,"spearman":0.35674470342177855,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.5612707655173639,"r2_score":-0.752005951422845,"spearman":0.05021647489380426,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":165.27548224596853,"r2_score":-753131.4060873595,"spearman":0.1977233957999042,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":1.153429251284502,"r2_score":-0.004781039811609489,"spearman":0.23193177120601463,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":206.1404016836662,"spearman":0.3086495056994413,"r2_score":-2354475.7131313877,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":114.21623075025657,"spearman":0.1675992316074245,"r2_score":-417650.6100248206,"validity":0.9918032786885246,"rank":4}],"name":"MGLUR5 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-nk1-ki":{"meta":{"title":"GPCR Affinity Suite: NK1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NK1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.1998305204785105,"r2_score":-0.21455117067464413,"spearman":0.21804443476113733,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.3255970723348225,"r2_score":-0.6309746148658673,"spearman":0.07248486964543523,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":53.121713939938445,"r2_score":-241911.80316447257,"spearman":0.19670836522564708,"validity":0.9988751406074241,"rank":6},{"model":"Grok 4.5, high","mae":1.2232617953469376,"r2_score":-0.2265455961098919,"spearman":0.08653726638963952,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":3.8619464906762846,"spearman":0.10714827042482518,"r2_score":-1954.1753642359379,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":4.53751656418195,"spearman":0.3077574080035947,"r2_score":-2621.11567242912,"validity":0.983127109111361,"rank":5}],"name":"NK1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-nk2-ic50":{"meta":{"title":"GPCR Affinity Suite: NK2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NK2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.461387620181478,"r2_score":0.7427705371914975,"spearman":0.7693485825844895,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1484737375664094,"r2_score":-0.5355204029326506,"spearman":0.27295719070413643,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":244.2413467464985,"r2_score":-1250769.9716845115,"spearman":0.3325509307543602,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.0430843935800376,"r2_score":0.0007265892723765655,"spearman":0.23585897754616397,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":37.806896882672746,"spearman":0.3846146495408537,"r2_score":-162589.52668365196,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.301783269334354,"spearman":0.19301236474974345,"r2_score":-1.7622305079395533,"validity":0.9969604863221885,"rank":4}],"name":"NK2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-nk2-ki":{"meta":{"title":"GPCR Affinity Suite: NK2 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NK2 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.2099375810774433,"r2_score":-0.07939720885859858,"spearman":0.4411016551539415,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.4974649898991366,"r2_score":-0.7488031971231792,"spearman":0.12210374427610171,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":146.81857588491155,"r2_score":-660478.893059953,"spearman":0.2791087440253008,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.3446569451396755,"r2_score":-0.2356312266108429,"spearman":0.10601106408777627,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":25.240756119337963,"spearman":0.29032201606251584,"r2_score":-109704.86533738683,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":3.8162181770459274,"spearman":0.2195347611938072,"r2_score":-1109.9650339129064,"validity":0.9882629107981221,"rank":4}],"name":"NK2 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-nk3-ic50":{"meta":{"title":"GPCR Affinity Suite: NK3 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NK3 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6049460851632635,"r2_score":0.37376312158779634,"spearman":0.5034503213135104,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":0.8702942649048614,"r2_score":-0.47281535927368123,"spearman":0.2581430550740724,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":48.09506103207257,"r2_score":-419259.6952133013,"spearman":0.3240014318843096,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.8123319172364541,"r2_score":-0.11320920374834831,"spearman":0.26596382996926765,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":47.863938392394104,"spearman":0.18082166783513542,"r2_score":-418436.57881380786,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.7353465588191817,"spearman":0.28993046085400037,"r2_score":-164.42211398949252,"validity":0.9914163090128756,"rank":4}],"name":"NK3 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-nk3-ki":{"meta":{"title":"GPCR Affinity Suite: NK3 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NK3 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8908794610536671,"r2_score":-0.2701737231578205,"spearman":0.3017683104332532,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","mae":1.0032871426820604,"r2_score":-0.6287839487765299,"spearman":0.0686735414423169,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":25.282508326749955,"r2_score":-240548.2937389702,"spearman":0.0465350552188163,"validity":0.998792270531401,"rank":6},{"model":"Grok 4.5, high","mae":0.8860862938695798,"r2_score":-0.2393710839244716,"spearman":0.09700082151952003,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mae":2.6481606868543355,"spearman":0.07530675125282381,"r2_score":-1729.2695598309713,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.2354819755771338,"spearman":0.16461332281111501,"r2_score":-39.0116147875137,"validity":0.998792270531401,"rank":4}],"name":"NK3 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-npy1r-ki":{"meta":{"title":"GPCR Affinity Suite: NPY1R KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NPY1R KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.6965194297311322,"r2_score":0.3011695477444839,"spearman":0.5832120201906239,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.9507645213389948,"r2_score":-5.490264128937188,"spearman":-0.021321515525930933,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":114.11192504490562,"r2_score":-755872.5641591839,"spearman":0.30563160003215595,"validity":0.9946524064171123,"rank":6},{"model":"Grok 4.5, high","mae":0.9338483772460474,"r2_score":-0.16257810385394622,"spearman":0.2651729114308765,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":3.1300540093186635,"spearman":0.1646463661762435,"r2_score":-334.91725306660993,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.755598397747292,"spearman":0.01584121552899298,"r2_score":-4.100679749664899,"validity":0.8663101604278075,"rank":3}],"name":"NPY1R KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-npy2r-ic50":{"meta":{"title":"GPCR Affinity Suite: NPY2R IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NPY2R IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.5867770312348438,"r2_score":0.3186504520030119,"spearman":0.5870814770492723,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0559787234938343,"r2_score":-1.230958049885548,"spearman":0.28435184550041753,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.0129404550248895,"r2_score":-0.962675274379988,"spearman":0.425063366336073,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":0.829068774126135,"r2_score":-0.08924179621797812,"spearman":0.24022543088559167,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":18.778531339073425,"spearman":0.36388427867787476,"r2_score":-79322.22786778686,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.3771889827267836,"spearman":0.2083334227114461,"r2_score":-2.851750878630841,"validity":0.996551724137931,"rank":5}],"name":"NPY2R IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-npy5r-ic50":{"meta":{"title":"GPCR Affinity Suite: NPY5R IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NPY5R IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8993324963409648,"r2_score":0.0983317766834475,"spearman":0.42405888841694717,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0381858853143215,"r2_score":-0.3009361056924815,"spearman":0.1864434075658879,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.9594266783077884,"r2_score":-0.05609180685143422,"spearman":0.3290704301954008,"validity":0.9974937343358395,"rank":2},{"model":"Grok 4.5, high","mae":1.0070654838275306,"r2_score":-0.13630728825240102,"spearman":0.16258683805392876,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":1.052516175143397,"spearman":0.28312542330044577,"r2_score":-0.3258298267852775,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.0945481981462961,"spearman":0.3001849280702729,"r2_score":-1.2408673016247969,"validity":1,"rank":6}],"name":"NPY5R IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-npy5r-ki":{"meta":{"title":"GPCR Affinity Suite: NPY5R KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR NPY5R KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9453880811186914,"r2_score":0.037762988755709825,"spearman":0.42854723615473356,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2485642277408466,"r2_score":-1.000997605493958,"spearman":0.2635972082672466,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":68.12807058934972,"r2_score":-449568.4467725804,"spearman":0.255698016326636,"validity":0.9867986798679867,"rank":6},{"model":"Grok 4.5, high","mae":1.0306961998333086,"r2_score":-0.13767873484426363,"spearman":0.20386993645730123,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":45.593645214846234,"spearman":0.19915015095126623,"r2_score":-243221.27954296087,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.4355404852159517,"spearman":0.06456163633890481,"r2_score":-1.8555734317552388,"validity":0.9966996699669967,"rank":4}],"name":"NPY5R KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-oprd1-ic50":{"meta":{"title":"GPCR Affinity Suite: OPRD1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OPRD1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9578021097695065,"r2_score":0.31672151780799906,"spearman":0.63904491587308,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.5562000079827127,"r2_score":-0.8652579597581507,"spearman":0.23118883871528498,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":191.54021343818317,"r2_score":-859641.3897299203,"spearman":0.529923667618789,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.4225148019375347,"r2_score":-0.2992505059405035,"spearman":0.09606517313799744,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":62.582284604427365,"spearman":0.4875233299669652,"r2_score":-221542.8523076744,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":7.373585144221272,"spearman":0.5124834600066634,"r2_score":-3340.5448466800935,"validity":1,"rank":4}],"name":"OPRD1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-oprd1-ki":{"meta":{"title":"GPCR Affinity Suite: OPRD1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OPRD1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9047814710012552,"r2_score":0.388953009281675,"spearman":0.6804823143067728,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.6777713470771864,"r2_score":-1.3774999132587458,"spearman":0.14774022343491222,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":191.0779061013869,"r2_score":-864001.6130256621,"spearman":0.5220231411477114,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.3060400854757064,"r2_score":-0.1514086024366159,"spearman":0.14785625074524633,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":102.0601647750979,"spearman":0.49192260906536445,"r2_score":-375680.4425428107,"validity":0.999,"rank":5},{"model":"Claude Opus 5, high","mae":2.631014941633361,"spearman":0.43229450351718185,"r2_score":-183.29395713301057,"validity":1,"rank":4}],"name":"OPRD1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-oprk1-ic50":{"meta":{"title":"GPCR Affinity Suite: OPRK1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OPRK1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.014037839917152,"r2_score":0.07051532707516961,"spearman":0.3931186089594693,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.5231405795604331,"r2_score":-1.8497874550129616,"spearman":0.09965027800049338,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":56.252921409290735,"r2_score":-331088.6473838323,"spearman":0.31954208009280943,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.1667847475751503,"r2_score":-0.25473293712677303,"spearman":0.015271460978516645,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":28.76991470784724,"spearman":0.1271251465014082,"r2_score":-166311.33730607267,"validity":0.9972602739726028,"rank":5},{"model":"Claude Opus 5, high","mae":1.3987681894362745,"spearman":0.20340506282211626,"r2_score":-1.3200633170970755,"validity":1,"rank":3}],"name":"OPRK1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-oprk1-ki":{"meta":{"title":"GPCR Affinity Suite: OPRK1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OPRK1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":1.1502550286546125,"r2_score":0.03771710557855801,"spearman":0.5513594133750098,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.7270511013882424,"r2_score":-1.604160407493195,"spearman":0.11648268189188192,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":137.61396954675598,"r2_score":-615321.0776448003,"spearman":0.43461765512866696,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.318088307640891,"r2_score":-0.18078279219931503,"spearman":0.07202266416891714,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":59.87372564509018,"spearman":0.38948484833225966,"r2_score":-216499.75993140365,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":3.8045842430863903,"spearman":0.3252138188727268,"r2_score":-1100.1537099548725,"validity":0.999203187250996,"rank":4}],"name":"OPRK1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-oprm1-ic50":{"meta":{"title":"GPCR Affinity Suite: OPRM1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OPRM1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9885963409316955,"r2_score":0.08705550416039942,"spearman":0.32959304660989386,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.599275392217575,"r2_score":-1.4230748109346543,"spearman":-0.01063214567333721,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":204.06426755088225,"r2_score":-1130068.7438320308,"spearman":0.2183050323812748,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.2368549941639193,"r2_score":-0.285382924827984,"spearman":0.048245473292775926,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":112.49259114838253,"spearman":0.15243137032587897,"r2_score":-592950.5631177556,"validity":0.9977728285077951,"rank":5},{"model":"Claude Opus 5, high","mae":1.9920802608104022,"spearman":0.06993448574284872,"r2_score":-7.935702294842924,"validity":0.9955456570155902,"rank":4}],"name":"OPRM1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-oprm1-ki":{"meta":{"title":"GPCR Affinity Suite: OPRM1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OPRM1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.9106026327499813,"r2_score":0.07153227829459652,"spearman":0.3580325302744261,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.3405094965301827,"r2_score":-1.0746851458615332,"spearman":0.06744054239345912,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":29.5469932913259,"r2_score":-166930.19343774262,"spearman":0.22548751062575484,"validity":0.9996041171813144,"rank":5},{"model":"Grok 4.5, high","mae":1.07792120546645,"r2_score":-0.19516122823106397,"spearman":0.06697150906347697,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":46.67140964757085,"spearman":0.241400450904236,"r2_score":-270669.42289246456,"validity":0.9996041171813144,"rank":6},{"model":"Claude Opus 5, high","mae":2.484014518905494,"spearman":0.21457212725387037,"r2_score":-1515.2772640327585,"validity":0.9996041171813144,"rank":4}],"name":"OPRM1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ox1-ic50":{"meta":{"title":"GPCR Affinity Suite: OX1 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OX1 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8852164596506311,"r2_score":-0.04082603975524668,"spearman":0.3463137567365967,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0826895276118567,"r2_score":-0.6462638699450836,"spearman":0.04366619819093271,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":1.0408794642052244,"r2_score":-0.5119932731743375,"spearman":0.11853171225495177,"validity":0.9980302035456337,"rank":3},{"model":"Grok 4.5, high","mae":0.973185770103258,"r2_score":-0.2669275983229793,"spearman":0.04297478609419255,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":8.400871142118623,"spearman":0.14171426921983613,"r2_score":-57875.090454699195,"validity":0.9993434011818779,"rank":6},{"model":"Claude Opus 5, high","mae":1.4398550512618593,"spearman":0.1609275518860358,"r2_score":-180.44816884781383,"validity":0.9993434011818779,"rank":5}],"name":"OX1 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ox1-ki":{"meta":{"title":"GPCR Affinity Suite: OX1 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OX1 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8509882372902088,"r2_score":0.0783212204703495,"spearman":0.4999003858985006,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0968647056031862,"r2_score":-0.5511085715390831,"spearman":0.2023318893847379,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":1.1958705075814182,"r2_score":-0.8112605973009277,"spearman":0.1325676428607738,"validity":0.9981167608286252,"rank":4},{"model":"Grok 4.5, high","mae":0.9898089961525929,"r2_score":-0.16452720447592695,"spearman":0.14917938474397457,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":42.89355278297645,"spearman":0.22453473982886274,"r2_score":-299999.1579472043,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":2.3134928915593536,"spearman":0.2267648676716986,"r2_score":-406.0796737130239,"validity":1,"rank":5}],"name":"OX1 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ox2-ic50":{"meta":{"title":"GPCR Affinity Suite: OX2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OX2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8243894075095505,"r2_score":-0.23040736002813667,"spearman":0.14720826555001354,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0166364427951795,"r2_score":-0.8033481500485391,"spearman":0.00877071319697853,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":0.9761916600368153,"r2_score":-0.6857246143953966,"spearman":0.04003995474878778,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":0.8698254003006525,"r2_score":-0.2714047219219369,"spearman":0.0055967833970577115,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.2535203235038044,"spearman":0.06376759185480045,"r2_score":-6.212633703511744,"validity":0.9994141769185706,"rank":6},{"model":"Claude Opus 5, high","mae":1.0444557395300293,"spearman":0.008255845433924176,"r2_score":-5.815107418150141,"validity":1,"rank":5}],"name":"OX2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-ox2-ki":{"meta":{"title":"GPCR Affinity Suite: OX2 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR OX2 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8865570616568295,"r2_score":-0.04958549213117491,"spearman":0.2875756194133384,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.13639922594111,"r2_score":-0.7720602827893237,"spearman":0.09728740957428218,"validity":0.9924357034795764,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":16.467402234527864,"r2_score":-131938.05345831596,"spearman":0.015727665730051483,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":0.9947797195500147,"r2_score":-0.3317222585905135,"spearman":0.034854601647168554,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":2.7022379909560033,"spearman":0.05661264121763366,"r2_score":-733.2955572780186,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":10.333651230757964,"spearman":0.028707113374710012,"r2_score":-34306.98805171505,"validity":0.9984871406959153,"rank":5}],"name":"OX2 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-thrombin-ic50":{"meta":{"title":"GPCR Affinity Suite: THROMBIN IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR THROMBIN IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.7401268670740832,"r2_score":0.0018277903873952228,"spearman":0.21706548073356097,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2332957608869182,"r2_score":-1.2145541698310942,"spearman":0.009552019948248515,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","mae":1.1470047132626742,"r2_score":-0.9527107154143406,"spearman":0.0031560451293080934,"validity":1,"rank":4},{"model":"Grok 4.5, high","mae":0.9828685562020377,"r2_score":-0.42598243422365334,"spearman":0.008228726631154816,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.233232462813992,"spearman":0.028477604284912833,"r2_score":-1.1375040549854822,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.1222894520838607,"spearman":-0.04948749822439584,"r2_score":-0.9980120611806631,"validity":1,"rank":3}],"name":"THROMBIN IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-v1a-ic50":{"meta":{"title":"GPCR Affinity Suite: V1A IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR V1A IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.936602234014636,"r2_score":-0.07147151802222274,"spearman":0.3964404287673131,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.1386888592823996,"r2_score":-0.6579254956409504,"spearman":0.17742633212660094,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":74.56428082352473,"r2_score":-505009.6413977986,"spearman":0.1273989488685535,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.0592270940069637,"r2_score":-0.3370628082270928,"spearman":0.07278573907643054,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":2.6021222803889037,"spearman":0.08111018782476795,"r2_score":-320.6260087186484,"validity":0.9982078853046595,"rank":5},{"model":"Claude Opus 5, high","mae":1.1569782660207346,"spearman":0.143550135539829,"r2_score":-0.7978881422467572,"validity":0.996415770609319,"rank":4}],"name":"V1A IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-v1a-ki":{"meta":{"title":"GPCR Affinity Suite: V1A KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR V1A KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8242835804694736,"r2_score":0.13924301443101417,"spearman":0.3512679747709451,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.0819810310925966,"r2_score":-0.46863724601019174,"spearman":0.0851852319310482,"validity":1,"rank":5},{"model":"Gemini 3.1 Pro, high","mae":0.9344535192896362,"r2_score":-0.21581922689202315,"spearman":0.2586499003517999,"validity":1,"rank":2},{"model":"Grok 4.5, high","mae":0.985735375074761,"r2_score":-0.1753851368929642,"spearman":0.10011437951634158,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":1.1321827831159361,"spearman":0.22288218771896467,"r2_score":-1.179957447730677,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":1.0411722275852184,"spearman":0.176272423787058,"r2_score":-2.136742627764201,"validity":0.9977973568281938,"rank":4}],"name":"V1A KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-v2-ic50":{"meta":{"title":"GPCR Affinity Suite: V2 IC50","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR V2 IC50 affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8172969189775168,"r2_score":0.20535928372226808,"spearman":0.513459681920468,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2202296461378055,"r2_score":-0.9057939634978713,"spearman":0.053875885235108764,"validity":0.773109243697479,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":57.15705743691018,"r2_score":-421545.59688874776,"spearman":0.19917346143711456,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.0975167685785034,"r2_score":-0.3297879662152212,"spearman":0.05919584078860855,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":1.4699158819399005,"spearman":0.20046928424666702,"r2_score":-32.15671929798895,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":1.2748895371879003,"spearman":0.15332879496468038,"r2_score":-1.2004364016826896,"validity":0.9859943977591037,"rank":4}],"name":"V2 IC50","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-gpcr-affinity-suite-v2-ki":{"meta":{"title":"GPCR Affinity Suite: V2 KI","category":"Medicinal Chemistry > GPCR Affinity Suite","description":"GPCR V2 KI affinity prediction (IC50 or Ki).","metric_types":[],"paper_link":"#","source_file":"gpcr_ci (1).json","original_category":"Affinity and Binding"},"radar_keys":["mae","spearman","r2_score","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":0.8862828577216998,"r2_score":0.18154582890918647,"spearman":0.6060359535881994,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mae":1.2375028904520815,"r2_score":-0.5905401985026533,"spearman":0.2614554099203653,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","mae":10.98889373024881,"r2_score":-6192.062137279003,"spearman":0.3810859515210836,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":1.014737578598802,"r2_score":0.0004058037668380754,"spearman":0.27598634398425814,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":2.3714261399103473,"spearman":0.37861551759449774,"r2_score":-102.20975270726797,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":0.938795886688073,"spearman":0.5398505205394327,"r2_score":0.007748341290150895,"validity":1,"rank":2}],"name":"V2 KI","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-medchem-holdout-ic50-prediction-internal-holdout":{"meta":{"title":"IC50 Prediction: IC50 Prediction (Internal Holdout)","category":"Medicinal Chemistry > IC50 Prediction","description":"Evaluates IC50 (half-maximal inhibitory concentration) prediction using the in-house experimental data","metric_types":[],"paper_link":"#","source_file":"medchem_holdout_ci.json","original_category":"Affinity and Binding"},"radar_keys":["medchem_mean_absolute_error","medchem_r2_score","spearman","pearson","validity"],"leaderboard_metric":"medchem_mean_absolute_error","results":[{"model":"GPT 5.5, medium","medchem_mean_absolute_error":1.789750972562733,"medchem_r2_score":-0.26888830033872146,"spearman":0.13355480234531955,"pearson":0.1540115946079018,"validity":1,"rank":2},{"model":"Grok 4.5, high","medchem_mean_absolute_error":1.7755380429440342,"medchem_r2_score":-0.15318078995153428,"spearman":0.11080950599295247,"pearson":0.11766427002479837,"validity":1,"rank":1}],"name":"IC50 Prediction (Internal Holdout)","metric_info":{"medchem_mean_absolute_error":{"label":"medchem_mean_absolute_error","shortLabel":"MEDCH","direction":"lower_better"},"medchem_r2_score":{"label":"medchem_r2_score","shortLabel":"MEDCH","direction":"higher_better"},"spearman":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"pearson":{"label":"Pearson correlation","shortLabel":"Prs","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-molecular-mechanics-energy-mmff-energy-prediction":{"meta":{"title":"Molecular Mechanics Energy: MMFF Energy Prediction","category":"Medicinal Chemistry > Molecular Mechanics Energy","description":"Predicting Molecular Mechanics (MM) energy for diverse 3D conformations. A computationally designed drug is useless if it instantly unfolds or repels itself in physical reality. Ensuring small molecules have low-energy conformations is a vital sanity check in AI-driven drug design. Chemistry42 platform was used to generate diverse conformations for drug-like molecules; models predict the resulting MM energy.","metric_types":[],"paper_link":"#","source_file":"mm_energy_ci.json","original_category":"Affinity and Binding"},"radar_keys":["r2_score","mae","spearman","pearson","validity"],"leaderboard_metric":"r2_score","results":[{"model":"GPT 5.5, medium","mae":35.205183733897414,"r2_score":0.025720626919336298,"spearman":0.6145262060207076,"pearson":0.23821551463770713,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mae":48.38727409742917,"r2_score":-0.04641588726348167,"spearman":0.5681674022586494,"pearson":0.2551217597252734,"validity":0.9735849056603774,"rank":5},{"model":"Gemini 3.1 Pro, high","mae":34.15443068474103,"r2_score":0.0762227568175875,"spearman":0.6399671577870083,"pearson":0.31879736059034464,"validity":1,"rank":2},{"model":"Grok 4.5, high","mae":35.05271670267684,"r2_score":0.008416093108233214,"spearman":0.5687488091473033,"pearson":0.25217125804417456,"validity":1,"rank":4},{"model":"Claude Opus 5, high","r2_score":-0.1928708959802572,"mae":58.21554420723894,"spearman":0.6561017313940115,"pearson":0.2681894144965156,"validity":0.9735849056603774,"rank":6},{"model":"GPT 5.6 Sol, high","r2_score":0.28436199745866,"mae":30.973632454117404,"spearman":0.6883702683425589,"pearson":0.5536669968834724,"validity":1,"rank":1}],"name":"MMFF Energy Prediction","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"pearson":{"label":"Pearson Correlation","shortLabel":"PEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-molecular-mechanics-energy-conformation-ranking":{"meta":{"title":"Molecular Mechanics Energy: Conformation ranking","category":"Medicinal Chemistry > Molecular Mechanics Energy","description":"conformation_ranking","metric_types":[],"paper_link":"#","source_file":"mm_energy_ci.json","original_category":"Affinity and Binding"},"radar_keys":["kendalltau_score","validity"],"leaderboard_metric":"kendalltau_score","results":[{"model":"GPT 5.5, medium","kendalltau_score":0.3086792452830189,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","kendalltau_score":0.19536679536679535,"validity":0.9773584905660377,"rank":6},{"model":"Gemini 3.1 Pro, high","kendalltau_score":0.3924528301886792,"validity":1,"rank":2},{"model":"Grok 4.5, high","kendalltau_score":0.4437735849056604,"validity":1,"rank":1},{"model":"Claude Opus 5, high","kendalltau_score":0.35252100840336137,"validity":0.8981132075471698,"rank":4},{"model":"GPT 5.6 Sol, high","kendalltau_score":0.37584905660377355,"validity":1,"rank":3}],"name":"Conformation ranking","metric_info":{"kendalltau_score":{"label":"kendalltau_score","shortLabel":"KENDA","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"kendalltau_score","min":-1,"max":1}},"med-mumo-multiple-property-optimization-bbbp-drd2-plogp":{"meta":{"title":"MuMO: Multiple Property Optimization: BBBP + DRD2 + pLogP","category":"Medicinal Chemistry > MuMO: Multiple Property Optimization","description":"bbbp+drd2+plogp","metric_types":[],"paper_link":"#","source_file":"mumo_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["SR","Sim","RI","valid_fraction"],"leaderboard_metric":"SR","results":[{"model":"GPT 5.5, medium","SR":83.2,"Sim":0.5281577880524175,"RI":2.471461615559877,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","SR":92.4,"Sim":0.4824130400001649,"RI":4.273720496373264,"valid_fraction":0.996,"rank":3},{"model":"Gemini 3.1 Pro, high","SR":53.400000000000006,"Sim":0.43627843238955993,"RI":4.258947464242296,"valid_fraction":0.908,"rank":6},{"model":"Grok 4.5, high","SR":93.60000000000001,"Sim":0.49680636130410777,"RI":4.052242028308801,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","SR":92.60000000000001,"Sim":0.4213185110355848,"RI":6.355954934884994,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","SR":78,"Sim":0.6110858216647377,"RI":2.090384113582361,"valid_fraction":1,"rank":5}],"name":"BBBP + DRD2 + pLogP","metric_info":{"SR":{"label":"SR","shortLabel":"SRM","direction":"higher_better"},"Sim":{"label":"Sim","shortLabel":"SIM","direction":"higher_better"},"RI":{"label":"RI","shortLabel":"RIM","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"SR","min":0,"max":100}},"med-mumo-multiple-property-optimization-bbbp-drd2-qed":{"meta":{"title":"MuMO: Multiple Property Optimization: BBBP + DRD2 + QED","category":"Medicinal Chemistry > MuMO: Multiple Property Optimization","description":"bbbp+drd2+qed","metric_types":[],"paper_link":"#","source_file":"mumo_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["SR","Sim","RI","valid_fraction"],"leaderboard_metric":"SR","results":[{"model":"GPT 5.5, medium","SR":91.60000000000001,"Sim":0.4467701534368654,"RI":7.623355327587721,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","SR":66,"Sim":0.4638351931472066,"RI":5.042279094210417,"valid_fraction":0.998,"rank":5},{"model":"Gemini 3.1 Pro, high","SR":58.4,"Sim":0.44797920478673503,"RI":6.0326968282688584,"valid_fraction":0.894,"rank":6},{"model":"Grok 4.5, high","SR":93.4,"Sim":0.4962528963390213,"RI":5.383957152784991,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","SR":89.2,"Sim":0.4226239961456946,"RI":9.403539767043124,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","SR":82.6,"Sim":0.5412623247940878,"RI":5.171932879576363,"valid_fraction":1,"rank":4}],"name":"BBBP + DRD2 + QED","metric_info":{"SR":{"label":"SR","shortLabel":"SRM","direction":"higher_better"},"Sim":{"label":"Sim","shortLabel":"SIM","direction":"higher_better"},"RI":{"label":"RI","shortLabel":"RIM","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"SR","min":0,"max":100}},"med-mumo-multiple-property-optimization-bbbp-plogp-qed":{"meta":{"title":"MuMO: Multiple Property Optimization: BBBP + pLogP + QED","category":"Medicinal Chemistry > MuMO: Multiple Property Optimization","description":"bbbp+plogp+qed","metric_types":[],"paper_link":"#","source_file":"mumo_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["SR","Sim","RI","valid_fraction"],"leaderboard_metric":"SR","results":[{"model":"GPT 5.5, medium","SR":92.60000000000001,"Sim":0.5217665851679602,"RI":1.4253630386853993,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","SR":98.6,"Sim":0.42691677214067164,"RI":1.9552226368944814,"valid_fraction":0.998,"rank":1},{"model":"Gemini 3.1 Pro, high","SR":75,"Sim":0.516557740604782,"RI":1.3952495776411626,"valid_fraction":0.88,"rank":6},{"model":"Grok 4.5, high","SR":98,"Sim":0.48337901683183904,"RI":1.6481253587608822,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","SR":98.2,"Sim":0.45493334693932874,"RI":1.890093757393313,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","SR":92.60000000000001,"Sim":0.606150925971486,"RI":1.245200821704212,"valid_fraction":1,"rank":4}],"name":"BBBP + pLogP + QED","metric_info":{"SR":{"label":"SR","shortLabel":"SRM","direction":"higher_better"},"Sim":{"label":"Sim","shortLabel":"SIM","direction":"higher_better"},"RI":{"label":"RI","shortLabel":"RIM","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"SR","min":0,"max":100}},"med-mumo-multiple-property-optimization-drd2-plogp-qed":{"meta":{"title":"MuMO: Multiple Property Optimization: DRD2 + pLogP + QED","category":"Medicinal Chemistry > MuMO: Multiple Property Optimization","description":"drd2+plogp+qed","metric_types":[],"paper_link":"#","source_file":"mumo_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["SR","Sim","RI","valid_fraction"],"leaderboard_metric":"SR","results":[{"model":"GPT 5.5, medium","SR":72.39999999999999,"Sim":0.4493505673588282,"RI":3.443993396244723,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","SR":58.4,"Sim":0.39273162330169525,"RI":3.2883368212500725,"valid_fraction":0.998,"rank":5},{"model":"Gemini 3.1 Pro, high","SR":33.6,"Sim":0.40668755551724295,"RI":3.4588653333712305,"valid_fraction":0.876,"rank":6},{"model":"Grok 4.5, high","SR":72.2,"Sim":0.44889418378390294,"RI":2.7885021742607394,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","SR":89.2,"Sim":0.3294856408255967,"RI":6.979388085837652,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","SR":63.800000000000004,"Sim":0.5231270643898409,"RI":2.1561326168612274,"valid_fraction":1,"rank":4}],"name":"DRD2 + pLogP + QED","metric_info":{"SR":{"label":"SR","shortLabel":"SRM","direction":"higher_better"},"Sim":{"label":"Sim","shortLabel":"SIM","direction":"higher_better"},"RI":{"label":"RI","shortLabel":"RIM","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"SR","min":0,"max":100}},"med-mumo-multiple-property-optimization-bbbp-drd2-plogp-qed":{"meta":{"title":"MuMO: Multiple Property Optimization: BBBP + DRD2 + pLogP + QED","category":"Medicinal Chemistry > MuMO: Multiple Property Optimization","description":"bbbp+drd2+plogp+qed","metric_types":[],"paper_link":"#","source_file":"mumo_ci (1).json","original_category":"Molecular Design & Optimization"},"radar_keys":["SR","Sim","RI","valid_fraction"],"leaderboard_metric":"SR","results":[{"model":"GPT 5.5, medium","SR":77,"Sim":0.35594472907611496,"RI":4.565436555766778,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","SR":48,"Sim":0.38954141050581015,"RI":3.7797536508565535,"valid_fraction":0.988,"rank":5},{"model":"Gemini 3.1 Pro, high","SR":31.4,"Sim":0.3364980469396145,"RI":5.30899363789077,"valid_fraction":0.896,"rank":6},{"model":"Grok 4.5, high","SR":63.6,"Sim":0.4307031137769269,"RI":3.732595643339764,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","SR":65.60000000000001,"Sim":0.3539176575969704,"RI":6.59551526821051,"valid_fraction":0.996,"rank":3},{"model":"GPT 5.6 Sol, high","SR":66.8,"Sim":0.39364124336037043,"RI":5.632387644830176,"valid_fraction":1,"rank":2}],"name":"BBBP + DRD2 + pLogP + QED","metric_info":{"SR":{"label":"SR","shortLabel":"SRM","direction":"higher_better"},"Sim":{"label":"Sim","shortLabel":"SIM","direction":"higher_better"},"RI":{"label":"RI","shortLabel":"RIM","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"SR","min":0,"max":100}},"med-pdbbind-restore-interactions-3d-interactions-restoration":{"meta":{"title":"PDBBind Restore Interactions: 3D Interactions Restoration","category":"Medicinal Chemistry > PDBBind Restore Interactions","description":"Evaluates the ability to restore 3D protein-ligand interactions from LP-PDBBind dataset. Measures the ratio of restored interactions, power of restored interactions, and fake interactions, along with validity fraction.","metric_types":[],"paper_link":"#","source_file":"pdbbind_restore_interactions_ci.json","original_category":"Affinity and Binding"},"radar_keys":["interactions_3d_restored_ratio","interactions_3d_power_restored_ratio","interactions_3d_fake_ratio","validity"],"leaderboard_metric":"interactions_3d_restored_ratio","results":[{"model":"GPT 5.5, medium","interactions_3d_restored_ratio":0.461133023518972,"interactions_3d_power_restored_ratio":0.44206716022747655,"interactions_3d_fake_ratio":0.27060321212610566,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","interactions_3d_restored_ratio":0.24430283671934952,"interactions_3d_power_restored_ratio":0.5301363615119596,"interactions_3d_fake_ratio":0.6593514080129146,"validity":0.8387799564270153,"rank":6},{"model":"Gemini 3.1 Pro, high","interactions_3d_restored_ratio":0.3881091000356678,"interactions_3d_power_restored_ratio":0.6428110489077687,"interactions_3d_fake_ratio":0.5625815158811297,"validity":0.971677559912854,"rank":3},{"model":"Grok 4.5, high","interactions_3d_restored_ratio":0.36009623965343335,"interactions_3d_power_restored_ratio":0.4974890041290088,"interactions_3d_fake_ratio":0.6615949106174229,"validity":1,"rank":5},{"model":"Claude Opus 5, high","interactions_3d_restored_ratio":0.37414358943571535,"interactions_3d_power_restored_ratio":0.4810469399599269,"interactions_3d_fake_ratio":0.5605630097214622,"validity":0.8169934640522876,"rank":4},{"model":"GPT 5.6 Sol, high","interactions_3d_restored_ratio":0.49760773859637675,"interactions_3d_power_restored_ratio":0.5201738922228621,"interactions_3d_fake_ratio":0.20923536046387206,"validity":1,"rank":1}],"name":"3D Interactions Restoration","metric_info":{"interactions_3d_restored_ratio":{"label":"3D Interactions Restored Ratio","shortLabel":"Restored","direction":"higher_better"},"interactions_3d_power_restored_ratio":{"label":"3D Interactions Power Restored Ratio","shortLabel":"Power","direction":"higher_better"},"interactions_3d_fake_ratio":{"label":"3D Interactions Fake Ratio","shortLabel":"Fake","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"interactions_3d_restored_ratio","min":0,"max":1}},"med-pkis2-kinase-inhibition-suite-pkis2-ret":{"meta":{"title":"PKIS2 Kinase Inhibition Suite: PKIS2 RET","category":"Medicinal Chemistry > PKIS2 Kinase Inhibition Suite","description":"Kinase inhibition prediction for RET (Published Kinase Inhibitor Set 2). Predicting how small molecules interact with the RET kinase pocket is essential for designing selective kinase inhibitors in cancer therapy.","metric_types":[],"paper_link":"#","source_file":"pkis2_kinases_ci.json","original_category":"Affinity and Binding"},"radar_keys":["mae","r2_score","spearman","pearson","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":31.721254716981132,"r2_score":-0.36200175539270973,"spearman":0.5135686021709425,"pearson":0.4898366749353222,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":23.89462264150943,"r2_score":-0.005888041132392496,"spearman":0.4749523620752865,"pearson":0.47362602842104723,"validity":1,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":26.642452830188677,"r2_score":-0.40523239798358524,"spearman":0.3817589093985721,"pearson":0.4305515134386643,"validity":1,"rank":3},{"model":"Grok 4.5, high","mae":25.354716981132075,"r2_score":0.0028935696093518315,"spearman":0.41419679369967694,"pearson":0.35110307252650713,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":34.63490566037736,"r2_score":-0.8120851754306306,"spearman":0.33975157416999047,"pearson":0.3307809164053158,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mae":27.939622641509434,"r2_score":-0.25520372315412,"spearman":0.4482701552777357,"pearson":0.3762921107151366,"validity":1,"rank":4}],"name":"PKIS2 RET","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"pearson":{"label":"Pearson Correlation","shortLabel":"PEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-pkis2-kinase-inhibition-suite-pkis2-kit":{"meta":{"title":"PKIS2 Kinase Inhibition Suite: PKIS2 KIT","category":"Medicinal Chemistry > PKIS2 Kinase Inhibition Suite","description":"Kinase inhibition prediction for KIT (Published Kinase Inhibitor Set 2). Essential for designing selective kinase inhibitors in cancer therapy.","metric_types":[],"paper_link":"#","source_file":"pkis2_kinases_ci.json","original_category":"Affinity and Binding"},"radar_keys":["mae","r2_score","spearman","pearson","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":28.885344827586206,"r2_score":-0.2186750806161497,"spearman":0.3429573171386667,"pearson":0.35045218579147636,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","mae":27.829310344827583,"r2_score":-0.21091135646026382,"spearman":0.35742535283246796,"pearson":0.4161907378881338,"validity":1,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":37.36810344827586,"r2_score":-1.1081281213843481,"spearman":0.2490235781703816,"pearson":0.2783032328307663,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":28.488793103448273,"r2_score":-0.1552570570198264,"spearman":0.2539701347308108,"pearson":0.29112483890427326,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":28.599999999999998,"r2_score":-0.2873525302762807,"spearman":0.3263157372640046,"pearson":0.3893920366436992,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":28.44051724137931,"r2_score":-0.2752893347967955,"spearman":0.40961405820796665,"pearson":0.4493930710975245,"validity":1,"rank":2}],"name":"PKIS2 KIT","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"pearson":{"label":"Pearson Correlation","shortLabel":"PEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-pkis2-kinase-inhibition-suite-pkis2-egfr":{"meta":{"title":"PKIS2 Kinase Inhibition Suite: PKIS2 EGFR","category":"Medicinal Chemistry > PKIS2 Kinase Inhibition Suite","description":"Kinase inhibition prediction for EGFR (Published Kinase Inhibitor Set 2). Essential for designing selective kinase inhibitors in cancer therapy.","metric_types":[],"paper_link":"#","source_file":"pkis2_kinases_ci.json","original_category":"Affinity and Binding"},"radar_keys":["mae","r2_score","spearman","pearson","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":23.130833333333335,"r2_score":-0.30974211126610873,"spearman":0.367249716987871,"pearson":0.4589075564650463,"validity":1,"rank":6},{"model":"Claude Opus 4.8, high","mae":19.87986111111111,"r2_score":-0.0476969621650003,"spearman":0.3665475414084238,"pearson":0.5672622963687317,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","mae":22.961111111111112,"r2_score":-0.3692882553628498,"spearman":0.33187268082368593,"pearson":0.5368443891320946,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":18.081249999999997,"r2_score":0.20098472248628063,"spearman":0.28629892760165315,"pearson":0.48538651650034215,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mae":21.196527777777778,"r2_score":-0.15118943561986153,"spearman":0.32645348810597985,"pearson":0.48677779330720133,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":19.619444444444444,"r2_score":-0.02293152516757213,"spearman":0.35229854989245124,"pearson":0.5979580360142911,"validity":1,"rank":2}],"name":"PKIS2 EGFR","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"},"pearson":{"label":"Pearson Correlation","shortLabel":"PEAR","direction":"higher_better"},"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-protacs-pk-bioavailability-f-upon-oral-administration-in-mice":{"meta":{"title":"ISM Benchmarks: PROTACs PK: Bioavailability F (%) upon oral administration in mice","category":"Medicinal Chemistry > ISM Benchmarks: PROTACs PK","description":"f_mice_pred_single","metric_types":[],"paper_link":"#","source_file":"protac_pk_ci.json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","r2_score","spearman","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":16.734666666666666,"r2_score":0.03930552458151759,"spearman":0.41690032647428144,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mae":18.508666666666667,"r2_score":-0.04692030898657973,"spearman":0.15859587356447594,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","mae":16.263333333333332,"r2_score":0.028420814775387715,"spearman":0.41852727172137816,"validity":1,"rank":2},{"model":"Grok 4.5, high","mae":17.293333333333333,"r2_score":-0.16639317053479852,"spearman":0.38101430420523175,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":17.053333333333335,"r2_score":0.0430730687052745,"spearman":0.3521913453920553,"validity":1,"rank":4},{"model":"GPT 5.6 Sol, high","mae":15.28,"r2_score":0.04562054673606608,"spearman":0.49944889362054595,"validity":1,"rank":1}],"name":"Bioavailability F (%) upon oral administration in mice","metric_info":{"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearman":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-protacs-pk-dnauc-dose-normalized-auc-0-t-ngmlhmg-upon-iv-administration-in-mice":{"meta":{"title":"ISM Benchmarks: PROTACs PK: DNAUC (dose-normalized AUC, 0-t) (ng/mL·h)/mg upon IV administration in mice","category":"Medicinal Chemistry > ISM Benchmarks: PROTACs PK","description":"dnauc_mice_pred_single","metric_types":[],"paper_link":"#","source_file":"protac_pk_ci.json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","r2_score","spearman","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":2762.093023255814,"r2_score":0.05022699368215511,"spearman":0.2825536253437017,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","mae":3415.75,"r2_score":-0.3485624672406522,"spearman":0.023836215832857947,"validity":0.9534883720930233,"rank":6},{"model":"Gemini 3.1 Pro, high","mae":3207.279069767442,"r2_score":-0.2381398514949995,"spearman":0.34080962321027947,"validity":1,"rank":5},{"model":"Grok 4.5, high","mae":2900.3488372093025,"r2_score":-0.13595287115098809,"spearman":0.182314145546216,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mae":2845.389534883721,"r2_score":-0.10800739483195465,"spearman":0.41666949728371816,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","mae":2726.4302325581393,"r2_score":-0.03622293498668716,"spearman":0.3609514944543805,"validity":1,"rank":1}],"name":"DNAUC (dose-normalized AUC, 0-t) (ng/mL·h)/mg upon IV administration in mice","metric_info":{"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearman":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-ism-benchmarks-protacs-pk-half-life-t12-h-upon-iv-administration-in-mice":{"meta":{"title":"ISM Benchmarks: PROTACs PK: Half-life T(1/2) (h) upon IV administration in mice","category":"Medicinal Chemistry > ISM Benchmarks: PROTACs PK","description":"t_half_mice_pred_single","metric_types":[],"paper_link":"#","source_file":"protac_pk_ci.json","original_category":"ADMET, PK & Safety"},"radar_keys":["mae","r2_score","spearman","validity"],"leaderboard_metric":"mae","results":[{"model":"GPT 5.5, medium","mae":3.0112499999999995,"r2_score":0.06594500091000788,"spearman":0.5983479103089996,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","mae":2.80125,"r2_score":0.10812648011498993,"spearman":0.7157460419812424,"validity":1,"rank":1},{"model":"Gemini 3.1 Pro, high","mae":3.1677500000000003,"r2_score":-0.0820388954192306,"spearman":0.6290434716533452,"validity":1,"rank":6},{"model":"Grok 4.5, high","mae":3.09875,"r2_score":-0.06362082973137295,"spearman":0.5551777446776308,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mae":2.8775,"r2_score":0.11408074879403696,"spearman":0.7023245996876724,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","mae":2.93125,"r2_score":0.08386878558244759,"spearman":0.6295987275849727,"validity":1,"rank":3}],"name":"Half-life T(1/2) (h) upon IV administration in mice","metric_info":{"mae":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearman":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-caco-2-permeability":{"meta":{"title":"TDC ADMET: Caco-2 Permeability","category":"Medicinal Chemistry > TDC ADMET","description":"Caco-2 cell effective permeability measures a compound's ability to cross the intestinal epithelial barrier via passive diffusion and active transport. Essential for predicting oral drug absorption. Dataset from Wang et al., 906 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mean_absolute_error","validity"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":0.3870471467032966,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","mean_absolute_error":0.4811482994505494,"validity":1,"rank":2},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":0.5360013423076924,"validity":1,"rank":5},{"model":"Grok 4.5, high","mean_absolute_error":0.5052243950549449,"validity":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.5135613631868131,"validity":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.5994238686813186,"validity":1,"rank":6}],"name":"Caco-2 Permeability","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-lipophilicity":{"meta":{"title":"TDC ADMET: Lipophilicity","category":"Medicinal Chemistry > TDC ADMET","description":"Experimental lipophilicity measured as octanol/water distribution coefficient (logD at pH 7.4). A key physicochemical property influencing absorption, distribution, metabolism, and toxicity. Dataset from AstraZeneca, 4,200 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mean_absolute_error","validity"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":0.7025357142857143,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7480625752105896,"validity":1,"rank":5},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":0.684470238095238,"validity":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.7281785714285713,"validity":1,"rank":4},{"model":"Claude Opus 5, high","mean_absolute_error":0.5649166666666666,"validity":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.8203511904761905,"validity":1,"rank":6}],"name":"Lipophilicity","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-solubility":{"meta":{"title":"TDC ADMET: Solubility","category":"Medicinal Chemistry > TDC ADMET","description":"Aqueous solubility prediction in log mol/L. Solubility directly impacts oral bioavailability and is a critical parameter in drug formulation and development. Dataset from AqSolDB, 9,982 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mean_absolute_error","validity"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":0.7114778874504757,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7883114351173796,"validity":1,"rank":5},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":0.7220929034267402,"validity":1,"rank":4},{"model":"Grok 4.5, high","mean_absolute_error":0.7920809663291938,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mean_absolute_error":0.6024582927107808,"validity":0.9939909864797196,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.6961246780648974,"validity":1,"rank":2}],"name":"Solubility","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-ppbr":{"meta":{"title":"TDC ADMET: PPBR","category":"Medicinal Chemistry > TDC ADMET","description":"Plasma protein binding rate (%) from AstraZeneca. Only the unbound fraction of a drug is pharmacologically active; high binding reduces free drug concentration and can affect efficacy and clearance. 1,797 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mean_absolute_error","validity"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":8.388425760286225,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mean_absolute_error":8.308150807899462,"validity":1,"rank":2},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":8.89,"validity":1,"rank":5},{"model":"Grok 4.5, high","mean_absolute_error":8.933846153846153,"validity":1,"rank":6},{"model":"Claude Opus 5, high","mean_absolute_error":7.9964845173041885,"validity":0.9821109123434705,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":8.592593917710197,"validity":1,"rank":4}],"name":"PPBR","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-vdss":{"meta":{"title":"TDC ADMET: VDss","category":"Medicinal Chemistry > TDC ADMET","description":"Volume of distribution at steady state (L/kg) measures how extensively a drug distributes from plasma into body tissues. A key pharmacokinetic parameter for determining dosing regimens. Dataset from Lombardo et al., 1,130 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["spearmanr_score","validity"],"leaderboard_metric":"spearmanr_score","results":[{"model":"GPT 5.5, medium","spearmanr_score":0.7412355355664441,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","spearmanr_score":0.6344557326078276,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","spearmanr_score":0.7898307243110346,"validity":1,"rank":1},{"model":"Grok 4.5, high","spearmanr_score":0.6825909071568449,"validity":1,"rank":5},{"model":"Claude Opus 5, high","spearmanr_score":0.7217727547952636,"validity":0.995575221238938,"rank":3},{"model":"GPT 5.6 Sol, high","spearmanr_score":0.6834376343204022,"validity":1,"rank":4}],"name":"VDss","metric_info":{"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-half-life":{"meta":{"title":"TDC ADMET: Half Life","category":"Medicinal Chemistry > TDC ADMET","description":"Drug half-life (hours) measures the time required for the plasma concentration to decrease by 50%. Critical for determining dosing frequency and achieving steady-state concentrations. Dataset from Obach et al., 667 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["spearmanr_score","validity"],"leaderboard_metric":"spearmanr_score","results":[{"model":"GPT 5.5, medium","spearmanr_score":0.8325884020421905,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","spearmanr_score":0.6927664280752739,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","spearmanr_score":0.8918348275140173,"validity":1,"rank":1},{"model":"Grok 4.5, high","spearmanr_score":0.8369873578950305,"validity":1,"rank":4},{"model":"Claude Opus 5, high","spearmanr_score":0.8678412708955776,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","spearmanr_score":0.8741997973714102,"validity":1,"rank":2}],"name":"Half Life","metric_info":{"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-clearance-microsome":{"meta":{"title":"TDC ADMET: Clearance Microsome","category":"Medicinal Chemistry > TDC ADMET","description":"Intrinsic clearance measured in liver microsomes (mL/min/g). Reflects the rate of drug metabolism by microsomal enzymes, primarily CYP450s. Dataset from AstraZeneca, 1,102 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["spearmanr_score","validity"],"leaderboard_metric":"spearmanr_score","results":[{"model":"GPT 5.5, medium","spearmanr_score":0.5033875550075814,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","spearmanr_score":0.39071776325780366,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","spearmanr_score":0.5788095475953313,"validity":1,"rank":1},{"model":"Grok 4.5, high","spearmanr_score":0.5009105726406252,"validity":1,"rank":4},{"model":"Claude Opus 5, high","spearmanr_score":0.552422680892007,"validity":0.995475113122172,"rank":2},{"model":"GPT 5.6 Sol, high","spearmanr_score":0.4756828377565498,"validity":1,"rank":5}],"name":"Clearance Microsome","metric_info":{"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-clearance-hepatocyte":{"meta":{"title":"TDC ADMET: Clearance Hepatocyte","category":"Medicinal Chemistry > TDC ADMET","description":"Intrinsic clearance measured in hepatocytes (µL/min/10^6 cells). Provides a more physiologically complete assessment of drug metabolism than microsomes, capturing both Phase I and Phase II metabolism. Dataset from AstraZeneca, 1,020 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["spearmanr_score","validity"],"leaderboard_metric":"spearmanr_score","results":[{"model":"GPT 5.5, medium","spearmanr_score":0.2922931462755926,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","spearmanr_score":0.21303999241669547,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","spearmanr_score":0.39370975570938505,"validity":1,"rank":1},{"model":"Grok 4.5, high","spearmanr_score":0.24254770353636984,"validity":1,"rank":5},{"model":"Claude Opus 5, high","spearmanr_score":0.3364263084088267,"validity":0.9917695473251029,"rank":2},{"model":"GPT 5.6 Sol, high","spearmanr_score":0.3255130192515649,"validity":1,"rank":3}],"name":"Clearance Hepatocyte","metric_info":{"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-p-gp-inhibition":{"meta":{"title":"TDC ADMET: P-gp Inhibition","category":"Medicinal Chemistry > TDC ADMET","description":"P-glycoprotein (P-gp) inhibition classification. P-gp is an efflux transporter that limits oral drug absorption and brain penetration. Inhibiting P-gp can alter drug pharmacokinetics and cause drug-drug interactions. Dataset from Broccatelli et al., 1,212 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7980205278592376,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.8293455078645695,"validity":1,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.7932551319648093,"validity":1,"rank":5},{"model":"Grok 4.5, high","balanced_accuracy_score":0.8194814716075713,"validity":1,"rank":3},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.8516129032258064,"validity":0.9959183673469387,"rank":1},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.7900559850706478,"validity":1,"rank":6}],"name":"P-gp Inhibition","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-hia":{"meta":{"title":"TDC ADMET: HIA","category":"Medicinal Chemistry > TDC ADMET","description":"Human intestinal absorption (HIA) classification predicting whether a drug is absorbed through the human intestine (>80% absorbed = positive). A critical first step in oral drug bioavailability. Dataset from Hou et al., 578 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.9592592592592593,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.9351851851851851,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.9537037037037037,"validity":1,"rank":3},{"model":"Grok 4.5, high","balanced_accuracy_score":0.9648148148148148,"validity":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.9537037037037037,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.9537037037037037,"validity":1,"rank":3}],"name":"HIA","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-bbb-penetration":{"meta":{"title":"TDC ADMET: BBB Penetration","category":"Medicinal Chemistry > TDC ADMET","description":"Blood-brain barrier penetration classification predicting whether a compound can cross the BBB. Essential for CNS drug design and for avoiding CNS-related toxicity in peripherally-targeted drugs. Dataset from Martins et al., 1,975 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.8483818011257036,"validity":1,"rank":5},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.8526884700665189,"validity":1,"rank":4},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.8663617886178863,"validity":1,"rank":2},{"model":"Grok 4.5, high","balanced_accuracy_score":0.8785959974984365,"validity":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.8287992495309568,"validity":1,"rank":6},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.8569418386491556,"validity":1,"rank":3}],"name":"BBB Penetration","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-bioavailability":{"meta":{"title":"TDC ADMET: Bioavailability","category":"Medicinal Chemistry > TDC ADMET","description":"Oral bioavailability classification predicting whether a drug achieves sufficient systemic exposure after oral administration (>20% bioavailable = positive). Dataset from Ma et al., 640 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.791652810109744,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.6326678765880218,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.791652810109744,"validity":1,"rank":3},{"model":"Grok 4.5, high","balanced_accuracy_score":0.8077818423678085,"validity":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7484203525108081,"validity":1,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.8026272031925508,"validity":1,"rank":2}],"name":"Bioavailability","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-cyp2c9-substrate":{"meta":{"title":"TDC ADMET: CYP2C9 Substrate","category":"Medicinal Chemistry > TDC ADMET","description":"CYP2C9 substrate classification predicting whether a compound is metabolized by the CYP2C9 enzyme. Important for predicting drug metabolism routes and potential drug-drug interactions. Dataset from CarbonMangels et al., 666 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.6852957135105806,"validity":1,"rank":1},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.5926478567552903,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.6424308193163321,"validity":1,"rank":5},{"model":"Grok 4.5, high","balanced_accuracy_score":0.6767498643516007,"validity":1,"rank":2},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.6481280520889854,"validity":1,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.6435160065111232,"validity":1,"rank":4}],"name":"CYP2C9 Substrate","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-cyp2c9-inhibition":{"meta":{"title":"TDC ADMET: CYP2C9 Inhibition","category":"Medicinal Chemistry > TDC ADMET","description":"CYP2C9 inhibition classification predicting whether a compound inhibits the CYP2C9 enzyme. CYP2C9 metabolizes ~15% of clinically used drugs; inhibition can cause dangerous drug-drug interactions. Dataset from Veith et al., 12,092 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7377694470477976,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7525047750709544,"validity":1,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.6560312034842053,"validity":1,"rank":6},{"model":"Grok 4.5, high","balanced_accuracy_score":0.682658911737141,"validity":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.76752463503284,"validity":0.9904919388176933,"rank":1},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.7335992691359581,"validity":1,"rank":4}],"name":"CYP2C9 Inhibition","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-cyp2d6-substrate":{"meta":{"title":"TDC ADMET: CYP2D6 Substrate","category":"Medicinal Chemistry > TDC ADMET","description":"CYP2D6 substrate classification predicting whether a compound is metabolized by the CYP2D6 enzyme. CYP2D6 is highly polymorphic, making substrate prediction critical for personalized medicine. Dataset from CarbonMangels et al., 664 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7836198179979778,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7283414260158446,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.7805864509605662,"validity":1,"rank":4},{"model":"Grok 4.5, high","balanced_accuracy_score":0.7665571284125379,"validity":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7875379170879677,"validity":1,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.7952477249747218,"validity":1,"rank":1}],"name":"CYP2D6 Substrate","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-cyp2d6-inhibition":{"meta":{"title":"TDC ADMET: CYP2D6 Inhibition","category":"Medicinal Chemistry > TDC ADMET","description":"CYP2D6 inhibition classification predicting whether a compound inhibits the CYP2D6 enzyme. CYP2D6 metabolizes ~25% of marketed drugs; inhibition can lead to severe adverse effects. Dataset from Veith et al., 13,130 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7352603588782186,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7443185973380138,"validity":1,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.6194071984522304,"validity":1,"rank":6},{"model":"Grok 4.5, high","balanced_accuracy_score":0.6420276112666211,"validity":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7376113909337466,"validity":0.9790555978674791,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.7511723568845656,"validity":1,"rank":1}],"name":"CYP2D6 Inhibition","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-cyp3a4-substrate":{"meta":{"title":"TDC ADMET: CYP3A4 Substrate","category":"Medicinal Chemistry > TDC ADMET","description":"CYP3A4 substrate classification predicting whether a compound is metabolized by the CYP3A4 enzyme. CYP3A4 is the most abundant hepatic CYP and metabolizes ~50% of marketed drugs. Dataset from CarbonMangels et al., 667 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7108951175406871,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.6354895104895104,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.6959764918625677,"validity":1,"rank":5},{"model":"Grok 4.5, high","balanced_accuracy_score":0.72875226039783,"validity":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.6993939393939395,"validity":0.9629629629629629,"rank":4},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.72875226039783,"validity":1,"rank":1}],"name":"CYP3A4 Substrate","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-cyp3a4-inhibition":{"meta":{"title":"TDC ADMET: CYP3A4 Inhibition","category":"Medicinal Chemistry > TDC ADMET","description":"CYP3A4 inhibition classification predicting whether a compound inhibits the CYP3A4 enzyme. Given CYP3A4's dominant role in drug metabolism, inhibition poses major drug-drug interaction risks. Dataset from Veith et al., 12,328 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.6954206343380689,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7014703086380633,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.7147383839260095,"validity":1,"rank":1},{"model":"Grok 4.5, high","balanced_accuracy_score":0.6830438351228171,"validity":1,"rank":6},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.708827154154257,"validity":0.9870287798946088,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.6914641958667263,"validity":1,"rank":5}],"name":"CYP3A4 Inhibition","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-ld50-acute-toxicity":{"meta":{"title":"TDC ADMET: LD50 Acute Toxicity","category":"Medicinal Chemistry > TDC ADMET","description":"Acute toxicity prediction as LD50 in log(1/(mol/kg)). LD50 is the dose required to kill 50% of test animals and is a fundamental safety metric in drug development. Dataset from Zhu et al., 7,385 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["mean_absolute_error","validity"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":0.5258443843031123,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","mean_absolute_error":0.6632369350282487,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":0.4109698917456022,"validity":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.5857604871447902,"validity":1,"rank":5},{"model":"Claude Opus 5, high","mean_absolute_error":0.5515439831104855,"validity":0.9614343707713126,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.48740866035182684,"validity":1,"rank":2}],"name":"LD50 Acute Toxicity","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-dili":{"meta":{"title":"TDC ADMET: DILI","category":"Medicinal Chemistry > TDC ADMET","description":"Drug-Induced Liver Injury (DILI) classification predicting whether a drug causes liver damage. DILI is a leading cause of drug withdrawals and clinical trial failures. 475 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.8421739130434782,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.5704347826086956,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.8447826086956521,"validity":1,"rank":3},{"model":"Grok 4.5, high","balanced_accuracy_score":0.8713043478260869,"validity":1,"rank":2},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.6951219512195121,"validity":0.8958333333333334,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.8730434782608696,"validity":1,"rank":1}],"name":"DILI","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-herg-blockers":{"meta":{"title":"TDC ADMET: hERG Blockers","category":"Medicinal Chemistry > TDC ADMET","description":"hERG channel blocker classification predicting whether a compound inhibits the human ether-à-go-go-related gene potassium channel. hERG inhibition can cause fatal cardiac arrhythmias (QT prolongation). Dataset from 648 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7849779086892489,"validity":1,"rank":3},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7611111111111111,"validity":1,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.76480117820324,"validity":1,"rank":4},{"model":"Grok 4.5, high","balanced_accuracy_score":0.7409425625920472,"validity":1,"rank":6},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.8092436974789916,"validity":0.9090909090909091,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.8135493372606775,"validity":1,"rank":1}],"name":"hERG Blockers","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-tdc-admet-ames-mutagenicity":{"meta":{"title":"TDC ADMET: AMES Mutagenicity","category":"Medicinal Chemistry > TDC ADMET","description":"Ames mutagenicity test classification predicting whether a compound is mutagenic. The Ames test is a regulatory requirement for drug candidates and uses bacterial reverse mutation to detect genotoxic potential. 7,255 compounds.","metric_types":[],"paper_link":"#","source_file":"tdc_ci (1).json","original_category":"ADMET, PK & Safety"},"radar_keys":["balanced_accuracy_score","validity"],"leaderboard_metric":"balanced_accuracy_score","results":[{"model":"GPT 5.5, medium","balanced_accuracy_score":0.7492852806986626,"validity":1,"rank":4},{"model":"Claude Opus 4.8, high","balanced_accuracy_score":0.7372564244383935,"validity":1,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy_score":0.7859112181558283,"validity":1,"rank":1},{"model":"Grok 4.5, high","balanced_accuracy_score":0.7437741095380759,"validity":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy_score":0.7756710795139177,"validity":0.997254632807138,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy_score":0.763509173862813,"validity":1,"rank":3}],"name":"AMES Mutagenicity","metric_info":{"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"validity":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"med-kinases-affinity-suite-abl1":{"meta":{"title":"Kinases Affinity Suite: ABL1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"abl1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0529077526494588,"r2_score":0.30639977450269273,"spearmanr_score":0.6021044131729759,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":1.253454171743407,"r2_score":0.28283637620042057,"spearmanr_score":0.5447424058391734,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":1.0997915505645777,"r2_score":0.22970727446007633,"spearmanr_score":0.6283640628519647,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.3101176979821973,"r2_score":-0.011558331111102715,"spearmanr_score":0.508926186316475,"valid_fraction":1,"rank":4}],"name":"ABL1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-akt1":{"meta":{"title":"Kinases Affinity Suite: AKT1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"akt1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.4401746562296283,"r2_score":-1.3516022964181502,"spearmanr_score":0.020831367088668215,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":1.1751287998931386,"r2_score":-0.11103253451694584,"spearmanr_score":0.17382811147563237,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.7711694590093954,"r2_score":-2.3569027945781555,"spearmanr_score":-0.05345291168528651,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.354610465908374,"r2_score":-0.803947705179723,"spearmanr_score":0.2429232338397035,"valid_fraction":1,"rank":2}],"name":"AKT1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-akt2":{"meta":{"title":"Kinases Affinity Suite: AKT2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"akt2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.2276147503743717,"r2_score":-0.322408197715512,"spearmanr_score":0.24231691678668743,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.1976964054749608,"r2_score":-0.08247747053256083,"spearmanr_score":0.14890222583177684,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.3664537586826864,"r2_score":-0.8142385108103478,"spearmanr_score":0.13119194487743274,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.6994282114459682,"r2_score":-2.293067805421474,"spearmanr_score":0.08025930773782117,"valid_fraction":1,"rank":4}],"name":"AKT2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-atm":{"meta":{"title":"Kinases Affinity Suite: ATM","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"atm_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9201346052511926,"r2_score":-0.1497209509070807,"spearmanr_score":0.4178549943547502,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.8133590357499827,"r2_score":0.16225803834269859,"spearmanr_score":0.45870151217270144,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.8115766523603328,"r2_score":0.07456835074659685,"spearmanr_score":0.45897904903894177,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0243011257154853,"r2_score":-0.3735082729700532,"spearmanr_score":0.42568939053100596,"valid_fraction":1,"rank":4}],"name":"ATM","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-atr":{"meta":{"title":"Kinases Affinity Suite: ATR","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"atr_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9190862302743472,"r2_score":-0.17207114827602243,"spearmanr_score":0.1134206571617879,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.8620953477969189,"r2_score":0.03331298073323219,"spearmanr_score":0.22219732017118554,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.8406915449897824,"r2_score":0.02328832641416334,"spearmanr_score":0.27719188089870633,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.2811989827919654,"r2_score":-3.5631558218219457,"spearmanr_score":0.2559401430680665,"valid_fraction":1,"rank":4}],"name":"ATR","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-aurka":{"meta":{"title":"Kinases Affinity Suite: AURKA","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"aurka_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.354685145242009,"r2_score":-0.05755780221653528,"spearmanr_score":0.3286479819790604,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","mean_absolute_error":1.3304277959188495,"r2_score":0.003978160467429825,"spearmanr_score":0.25169316690927585,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":1.2723612956929604,"r2_score":-0.024379750217681906,"spearmanr_score":0.35743218596912607,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.3312334162225772,"r2_score":-0.2735484014708094,"spearmanr_score":0.3848851399392336,"valid_fraction":1,"rank":3}],"name":"AURKA","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-aurkb":{"meta":{"title":"Kinases Affinity Suite: AURKB","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"aurkb_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.3521430723037957,"r2_score":-0.540265419406784,"spearmanr_score":0.07114451018291452,"valid_fraction":0.9754672897196262,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.2085120924386277,"r2_score":-0.10441823056168853,"spearmanr_score":0.10761010003277703,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.3991184847990643,"r2_score":-0.6110842840657715,"spearmanr_score":0.0339410221135112,"valid_fraction":0.9789719626168224,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":6.710912564226221,"r2_score":-9860.978632156162,"spearmanr_score":0.14803635821825528,"valid_fraction":1,"rank":4}],"name":"AURKB","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-axl":{"meta":{"title":"Kinases Affinity Suite: AXL","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"axl_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7777403548658938,"r2_score":0.04535495956959912,"spearmanr_score":0.46500620747268073,"valid_fraction":0.9557416267942583,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.8063019648620339,"r2_score":0.01987817203329001,"spearmanr_score":0.3234164707328867,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.7680294655348492,"r2_score":0.002224723407968754,"spearmanr_score":0.553897930043762,"valid_fraction":0.9641148325358851,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.197932308972426,"r2_score":-4.834175722610553,"spearmanr_score":0.3991669081517853,"valid_fraction":1,"rank":4}],"name":"AXL","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-braf":{"meta":{"title":"Kinases Affinity Suite: BRAF","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"braf_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.1880067178430118,"r2_score":-0.4560795956390644,"spearmanr_score":0.2089277620643178,"valid_fraction":0.9856819894498869,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.1006217373287162,"r2_score":-0.09839547181636266,"spearmanr_score":0.16383773211963606,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.2806454210657119,"r2_score":-0.774449397130506,"spearmanr_score":0.3247045485375104,"valid_fraction":0.9947249434815373,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":8.965194825696464,"r2_score":-45551.06379342807,"spearmanr_score":0.30615437232772924,"valid_fraction":1,"rank":4}],"name":"BRAF","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-btk":{"meta":{"title":"Kinases Affinity Suite: BTK","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"btk_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0685126435870986,"r2_score":-0.0889551788960583,"spearmanr_score":0.22543005224668056,"valid_fraction":0.9774070212026417,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.0476631939214958,"r2_score":-0.029027802368873923,"spearmanr_score":0.17289126572987898,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1011338946931508,"r2_score":-0.8690711730490321,"spearmanr_score":0.3476907903545905,"valid_fraction":0.9867917970107751,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":4.745034768605313,"r2_score":-22051.753288030428,"spearmanr_score":0.1940643992005482,"valid_fraction":1,"rank":4}],"name":"BTK","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-bub1":{"meta":{"title":"Kinases Affinity Suite: BUB1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"bub1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9130809118157455,"r2_score":0.009280638811132724,"spearmanr_score":0.2990576330853399,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.9279523961468116,"r2_score":0.17832603230838362,"spearmanr_score":0.36081174944650296,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.9186524682247149,"r2_score":-0.052902282868646955,"spearmanr_score":0.29378371306407325,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.1715829600896828,"r2_score":-0.522670564647955,"spearmanr_score":0.2091716067983624,"valid_fraction":1,"rank":4}],"name":"BUB1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-c-met":{"meta":{"title":"Kinases Affinity Suite: C-MET","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"c-met_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0585500737644844,"r2_score":-0.43696192634238273,"spearmanr_score":0.141626656149967,"valid_fraction":0.9892294946147473,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9167834307728298,"r2_score":-0.03980367330514101,"spearmanr_score":0.20537146121395997,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.2204878488306588,"r2_score":-0.8293979384418071,"spearmanr_score":0.3332099834998334,"valid_fraction":0.9983429991714996,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":13.640416790339383,"r2_score":-76888.85891052493,"spearmanr_score":0.22948636095515063,"valid_fraction":1,"rank":4}],"name":"C-MET","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-c-src":{"meta":{"title":"Kinases Affinity Suite: C-SRC","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"c-src_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.1890148104199603,"r2_score":-0.2897972175327572,"spearmanr_score":0.07703480418468481,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","mean_absolute_error":1.0955332833378895,"r2_score":-0.062489774329111514,"spearmanr_score":0.13989045269485373,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1354217087712166,"r2_score":-0.28050663967363065,"spearmanr_score":0.22095905218473222,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.169139418368018,"r2_score":-0.294231108079972,"spearmanr_score":0.24272597045729977,"valid_fraction":1,"rank":3}],"name":"C-SRC","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-cdk2":{"meta":{"title":"Kinases Affinity Suite: CDK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"cdk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9083636580471687,"r2_score":-0.2740072579195427,"spearmanr_score":0.20914962846220564,"valid_fraction":0.9936708860759493,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.8797743308200054,"r2_score":-0.011087806063616812,"spearmanr_score":0.25958664045893426,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9677530129168901,"r2_score":-0.5613752034232944,"spearmanr_score":0.32034428805110776,"valid_fraction":0.9981916817359855,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.1555477751838832,"r2_score":-2.2513853044737875,"spearmanr_score":0.32088125763655095,"valid_fraction":1,"rank":4}],"name":"CDK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-cdk7":{"meta":{"title":"Kinases Affinity Suite: CDK7","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"cdk7_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.02414546162717,"r2_score":-0.7578422780539846,"spearmanr_score":-0.07517958464510985,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.842707752146229,"r2_score":-0.09734268066876717,"spearmanr_score":0.1389570551537327,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.8731106482149319,"r2_score":-0.2486508906768452,"spearmanr_score":0.30519510808016986,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":6.661742614915492,"r2_score":-1336.2554862656084,"spearmanr_score":0.21583497798572718,"valid_fraction":1,"rank":4}],"name":"CDK7","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-cdk9":{"meta":{"title":"Kinases Affinity Suite: CDK9","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"cdk9_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0145782537616999,"r2_score":-0.36559541751109625,"spearmanr_score":0.2118228580669983,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.974661998424019,"r2_score":-0.07015045057716418,"spearmanr_score":0.14220754701696103,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.0913608022427617,"r2_score":-0.5843277734665533,"spearmanr_score":0.23766036152796263,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.4604572324863025,"r2_score":-1.929103052527771,"spearmanr_score":0.07485701655380539,"valid_fraction":1,"rank":4}],"name":"CDK9","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-chek1":{"meta":{"title":"Kinases Affinity Suite: CHEK1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"chek1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.8698111258317291,"r2_score":-1.1555974832018383,"spearmanr_score":0.08487835053895518,"valid_fraction":0.9969183359013868,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.479399678447279,"r2_score":-0.14554615486958355,"spearmanr_score":0.1005877700299696,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":2.2593443497274346,"r2_score":-2.2282895419362463,"spearmanr_score":0.077784418907837,"valid_fraction":0.99768875192604,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":2.934145768017394,"r2_score":-441.1328935824719,"spearmanr_score":0.24322367173853193,"valid_fraction":1,"rank":4}],"name":"CHEK1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-csnk1d":{"meta":{"title":"Kinases Affinity Suite: CSNK1D","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"csnk1d_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8845685086265087,"r2_score":-0.025406553202585114,"spearmanr_score":0.28778627845761506,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.9505327137747612,"r2_score":-0.10735128264018878,"spearmanr_score":0.17297438731912268,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","mean_absolute_error":0.8851588924145058,"r2_score":-0.15552172022069888,"spearmanr_score":0.2866578985781341,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.8960336596588818,"r2_score":-0.27881036825598704,"spearmanr_score":0.3511247169823634,"valid_fraction":1,"rank":3}],"name":"CSNK1D","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-csf1r":{"meta":{"title":"Kinases Affinity Suite: CSF1R","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"csf1r_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9700450134125478,"r2_score":-0.10661413019452781,"spearmanr_score":0.22637395950359823,"valid_fraction":0.9925373134328358,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":1.0059067180457435,"r2_score":-0.1307798709945014,"spearmanr_score":0.070138857343268,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":1.0487333033454018,"r2_score":-0.2771512699659042,"spearmanr_score":0.10003586803583132,"valid_fraction":0.9941956882255389,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.303675736388755,"r2_score":-1.7022554890487251,"spearmanr_score":0.02830172574794107,"valid_fraction":1,"rank":4}],"name":"CSF1R","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-ddr1":{"meta":{"title":"Kinases Affinity Suite: DDR1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"ddr1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7757716612318343,"r2_score":-0.2666752361580258,"spearmanr_score":0.06483224020905215,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.7222739654580006,"r2_score":-0.017138647704782795,"spearmanr_score":0.18115942288557507,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.7007985699005013,"r2_score":-0.08361849865006477,"spearmanr_score":0.27610608878564263,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.7774315118769589,"r2_score":-0.2696589640517353,"spearmanr_score":0.22270739801800507,"valid_fraction":1,"rank":4}],"name":"DDR1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-egrf":{"meta":{"title":"Kinases Affinity Suite: EGRF","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"egrf_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.36904180286349,"r2_score":-0.29857849498182465,"spearmanr_score":0.33938220729758545,"valid_fraction":0.9919549477071601,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.3602427208756178,"r2_score":-0.048985931163398666,"spearmanr_score":0.270870285544939,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.4156252413476011,"r2_score":-0.5276925455282693,"spearmanr_score":0.47766143355581453,"valid_fraction":0.996781979082864,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":11.717273894591894,"r2_score":-33383.637952090496,"spearmanr_score":0.40535574420569503,"valid_fraction":1,"rank":4}],"name":"EGRF","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-erbb2":{"meta":{"title":"Kinases Affinity Suite: ERBB2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"erbb2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.1816384114970873,"r2_score":-0.2689498851745171,"spearmanr_score":0.3274598630276498,"valid_fraction":0.9938042131350682,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.1760773338657236,"r2_score":-0.014694201584906308,"spearmanr_score":0.2694166978547563,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.4661399708988498,"r2_score":-81.13398123566601,"spearmanr_score":0.46496278226484394,"valid_fraction":0.9956629491945477,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":13.61061622863236,"r2_score":-64755.34452042307,"spearmanr_score":0.42986501563721075,"valid_fraction":1,"rank":4}],"name":"ERBB2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-erbb4":{"meta":{"title":"Kinases Affinity Suite: ERBB4","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"erbb4_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8600656620761937,"r2_score":0.25190526378230105,"spearmanr_score":0.5972108582205313,"valid_fraction":0.985981308411215,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.9863724543510148,"r2_score":0.22944484041766622,"spearmanr_score":0.48961267375240447,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.997351018787533,"r2_score":-0.19689773560920876,"spearmanr_score":0.5161273130420129,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0365283780208145,"r2_score":-0.17557578419970854,"spearmanr_score":0.5108703029157086,"valid_fraction":1,"rank":4}],"name":"ERBB4","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-fgfr2":{"meta":{"title":"Kinases Affinity Suite: FGFR2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"fgfr2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7340417841964627,"r2_score":-0.10596096147022682,"spearmanr_score":0.22714218178597087,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.7548412702511406,"r2_score":-0.13696940114155387,"spearmanr_score":0.16604761640317978,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.7781603154782628,"r2_score":-0.5619729860912583,"spearmanr_score":0.2699133849617668,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":20.963136040544626,"r2_score":-174656.20216441873,"spearmanr_score":0.1543717423937058,"valid_fraction":1,"rank":4}],"name":"FGFR2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-fgfr3":{"meta":{"title":"Kinases Affinity Suite: FGFR3","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"fgfr3_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7290457864778389,"r2_score":-0.15629434667381892,"spearmanr_score":0.17843317217437774,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.7123182580761385,"r2_score":-0.09246888112612028,"spearmanr_score":0.14777791935394463,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.719646754154154,"r2_score":-0.21080593863541908,"spearmanr_score":0.21996569375223582,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":2.6389653850383663,"r2_score":-1598.3903759453672,"spearmanr_score":0.1966044246099583,"valid_fraction":1,"rank":4}],"name":"FGFR3","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-flt3":{"meta":{"title":"Kinases Affinity Suite: FLT3","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"flt3_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.101907417132625,"r2_score":-0.13245749236777282,"spearmanr_score":0.3909754634368781,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":1.1647058284763618,"r2_score":0.03837982721965416,"spearmanr_score":0.3637994857663556,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":1.1299755854339857,"r2_score":-0.3176533964955879,"spearmanr_score":0.4887303865843872,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":4.330245048127535,"r2_score":-1145.2144168961895,"spearmanr_score":0.4261037219738396,"valid_fraction":1,"rank":4}],"name":"FLT3","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-gsk3b":{"meta":{"title":"Kinases Affinity Suite: GSK3B","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"gsk3b_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.1864846842255992,"r2_score":-0.6584172631442586,"spearmanr_score":0.055090736378818186,"valid_fraction":0.9846435100548446,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.0048788964977213,"r2_score":-0.15209210941624107,"spearmanr_score":0.05538784988943428,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.4225470510239824,"r2_score":-1.6037561388371442,"spearmanr_score":0.13202311412987797,"valid_fraction":0.9663619744058501,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":16.142570253249023,"r2_score":-427722.0782527006,"spearmanr_score":0.11770943421503773,"valid_fraction":1,"rank":4}],"name":"GSK3B","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-igf1r":{"meta":{"title":"Kinases Affinity Suite: IGF1R","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"igf1r_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.2295432605646064,"r2_score":0.05582598936279792,"spearmanr_score":0.4551437261879097,"valid_fraction":0.9603729603729604,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.350899730557017,"r2_score":0.12506266190638826,"spearmanr_score":0.3763029301441585,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":1.1966539824626867,"r2_score":0.10602033963229374,"spearmanr_score":0.49769594855436355,"valid_fraction":0.9627039627039627,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.5425054097142237,"r2_score":-7.606125600337856,"spearmanr_score":0.5119674192129013,"valid_fraction":1,"rank":4}],"name":"IGF1R","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-ikbkb":{"meta":{"title":"Kinases Affinity Suite: IKBKB","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"ikbkb_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8178077208933333,"r2_score":-0.9135016495084072,"spearmanr_score":0.10635852520803533,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.6973131236414605,"r2_score":-0.13487083654517296,"spearmanr_score":0.16270441968006982,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1884517865228088,"r2_score":-3.4117874164239863,"spearmanr_score":0.030090230385355573,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0428027247349965,"r2_score":-6.921126265972927,"spearmanr_score":0.20218290642124692,"valid_fraction":1,"rank":3}],"name":"IKBKB","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-insr":{"meta":{"title":"Kinases Affinity Suite: INSR","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"insr_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7136300390753242,"r2_score":0.06742419321329918,"spearmanr_score":0.5604405089482092,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.7865931547738576,"r2_score":0.14980961646520918,"spearmanr_score":0.5526271506573152,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.7648965574875295,"r2_score":0.007303841350422724,"spearmanr_score":0.49420366049746534,"valid_fraction":0.9844559585492227,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0487440456806407,"r2_score":-12.586668846863416,"spearmanr_score":0.5968375378207553,"valid_fraction":1,"rank":4}],"name":"INSR","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-irak4":{"meta":{"title":"Kinases Affinity Suite: IRAK4","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"irak4_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8703004022077127,"r2_score":-0.058656001469870045,"spearmanr_score":0.2856070119694688,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.8759400320769364,"r2_score":-0.027247649094853443,"spearmanr_score":0.22374563209625442,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.8942093699994503,"r2_score":-0.18793998776864074,"spearmanr_score":0.4080547862855247,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0679279214038877,"r2_score":-0.6817791304179421,"spearmanr_score":0.21866124735818768,"valid_fraction":1,"rank":4}],"name":"IRAK4","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-jak1":{"meta":{"title":"Kinases Affinity Suite: JAK1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"jak1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.1109387048942427,"r2_score":-0.24431572178075522,"spearmanr_score":0.1200709545955417,"valid_fraction":0.9703296703296703,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.0404133885188154,"r2_score":-0.0731500022987337,"spearmanr_score":0.11567719260903982,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1231595123997975,"r2_score":-0.31742992320366104,"spearmanr_score":0.23217861806531626,"valid_fraction":0.9917582417582418,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":6.749282271889726,"r2_score":-35456.714526473654,"spearmanr_score":0.2609210876308161,"valid_fraction":1,"rank":4}],"name":"JAK1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-jak2":{"meta":{"title":"Kinases Affinity Suite: JAK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"jak2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9903646692304854,"r2_score":-0.3891204865936746,"spearmanr_score":0.030066135567667435,"valid_fraction":0.9985401459854014,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.9051336405822782,"r2_score":-0.1278454481078659,"spearmanr_score":0.07443805262902348,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9256640317330337,"r2_score":-0.22759087690069624,"spearmanr_score":0.19149488737482254,"valid_fraction":0.9978102189781022,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.1965007515727708,"r2_score":-2.961896886953958,"spearmanr_score":0.1282282782090958,"valid_fraction":1,"rank":4}],"name":"JAK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-jak3":{"meta":{"title":"Kinases Affinity Suite: JAK3","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"jak3_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.2454897950777446,"r2_score":-0.41577719078029274,"spearmanr_score":0.10916138393998598,"valid_fraction":0.9877026271660145,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":1.0835867825913454,"r2_score":-0.057516375087071525,"spearmanr_score":0.11753685572403308,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1951014292732056,"r2_score":-0.6144861419008232,"spearmanr_score":0.16810703909963423,"valid_fraction":0.9932923420905534,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":6.891850394936314,"r2_score":-33724.9246641196,"spearmanr_score":0.25937380528826787,"valid_fraction":1,"rank":4}],"name":"JAK3","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-kit":{"meta":{"title":"Kinases Affinity Suite: KIT","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"kit_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9634747370358984,"r2_score":-0.31531652623540185,"spearmanr_score":0.23148754968985855,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.8839725940329252,"r2_score":-0.02790673113718878,"spearmanr_score":0.19998782419557573,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9787256731850461,"r2_score":-0.35700108930710805,"spearmanr_score":0.24591851331911424,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.181989165328997,"r2_score":-0.80571763416845,"spearmanr_score":0.061198340829610794,"valid_fraction":1,"rank":4}],"name":"KIT","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-lck":{"meta":{"title":"Kinases Affinity Suite: LCK","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"lck_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.2485441846854306,"r2_score":-0.4861893031467359,"spearmanr_score":0.1865556176979568,"valid_fraction":0.9946236559139785,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.2170411165936468,"r2_score":-0.16048062650040973,"spearmanr_score":0.07420637667795783,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.3257387310749842,"r2_score":-0.7385503383136565,"spearmanr_score":0.14956969713009824,"valid_fraction":0.9932795698924731,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.5462267525858904,"r2_score":-8.082385377955557,"spearmanr_score":0.07375694954250323,"valid_fraction":0.9993279569892473,"rank":4}],"name":"LCK","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-map2k1":{"meta":{"title":"Kinases Affinity Suite: MAP2K1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"map2k1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0953790566329873,"r2_score":-0.3336719613317023,"spearmanr_score":0.30862028110266065,"valid_fraction":0.9962121212121212,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9503607240526106,"r2_score":0.06413530493076658,"spearmanr_score":0.31053981056732555,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.214075140674849,"r2_score":-0.8597806740749818,"spearmanr_score":0.26405824881040574,"valid_fraction":0.9204545454545454,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":50.01681145206892,"r2_score":-264120.99791168846,"spearmanr_score":0.22642840944983142,"valid_fraction":1,"rank":4}],"name":"MAP2K1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-map4k4":{"meta":{"title":"Kinases Affinity Suite: MAP4K4","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"map4k4_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7967712559949262,"r2_score":-0.012272094798470023,"spearmanr_score":0.3099231125047132,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.8170596154747833,"r2_score":-0.05001835015000644,"spearmanr_score":0.174108776137888,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.7519689113168899,"r2_score":0.03704042077307046,"spearmanr_score":0.37632755539295426,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.109149542961496,"r2_score":-3.9980010353630293,"spearmanr_score":0.20752739144512816,"valid_fraction":1,"rank":4}],"name":"MAP4K4","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mapk10":{"meta":{"title":"Kinases Affinity Suite: MAPK10","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mapk10_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7648087691229655,"r2_score":-0.04310118756837844,"spearmanr_score":0.41679443433832336,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.7722823071723592,"r2_score":0.060793849468570915,"spearmanr_score":0.275402015518379,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.9176469972695297,"r2_score":-0.9006956840521518,"spearmanr_score":0.33086344717085814,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.8743088770372601,"r2_score":-0.2160291158190224,"spearmanr_score":0.44192413631461486,"valid_fraction":1,"rank":3}],"name":"MAPK10","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mapk11":{"meta":{"title":"Kinases Affinity Suite: MAPK11","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mapk11_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8124478203339064,"r2_score":-0.556362741273859,"spearmanr_score":0.16580259135657646,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.7882670183123143,"r2_score":-0.21814900194491837,"spearmanr_score":0.12847071973504287,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.8270067475077493,"r2_score":-1.237361140564206,"spearmanr_score":0.15580983801592796,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.7369447869994398,"r2_score":-0.2693317717777324,"spearmanr_score":0.3107777634481299,"valid_fraction":1,"rank":1}],"name":"MAPK11","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mapk14":{"meta":{"title":"Kinases Affinity Suite: MAPK14","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mapk14_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.2301809612350736,"r2_score":-0.340171599839562,"spearmanr_score":0.230158837684607,"valid_fraction":0.9939759036144579,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.115479483848534,"r2_score":-0.03235053231503615,"spearmanr_score":0.17563385304737694,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.407739337998943,"r2_score":-0.918526236290037,"spearmanr_score":0.24467973207266477,"valid_fraction":0.9951807228915662,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.4310709812649445,"r2_score":-0.734792657927448,"spearmanr_score":0.2448558924307603,"valid_fraction":0.9993975903614458,"rank":4}],"name":"MAPK14","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mapk8":{"meta":{"title":"Kinases Affinity Suite: MAPK8","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mapk8_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8025038538682118,"r2_score":-0.2226643915461397,"spearmanr_score":0.39042504098999603,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.7631491507547256,"r2_score":0.11234238602179525,"spearmanr_score":0.3531426746974052,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.0000162084045077,"r2_score":-0.7287082914683294,"spearmanr_score":0.22542940277206205,"valid_fraction":0.8300653594771242,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.8979791358202015,"r2_score":-0.45796762992443596,"spearmanr_score":0.3524745789557491,"valid_fraction":1,"rank":3}],"name":"MAPK8","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mapk9":{"meta":{"title":"Kinases Affinity Suite: MAPK9","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mapk9_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.5410673933738147,"r2_score":-0.09674874979479653,"spearmanr_score":0.5157242495614519,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.5153420176558691,"r2_score":0.2912276650513924,"spearmanr_score":0.5502395225925627,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.6952553632709811,"r2_score":-0.8477438549242053,"spearmanr_score":0.33008571144025567,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.713021555630571,"r2_score":-0.4812395240242784,"spearmanr_score":0.41178742743138413,"valid_fraction":1,"rank":4}],"name":"MAPK9","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mapkapk2":{"meta":{"title":"Kinases Affinity Suite: MAPKAPK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mapkapk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.3123706107092579,"r2_score":-1.1693548575003638,"spearmanr_score":0.2173906427366755,"valid_fraction":0.9966777408637874,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9804475910088878,"r2_score":0.006679803403887963,"spearmanr_score":0.22940250502579326,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.6120710760875998,"r2_score":-2.062597841957819,"spearmanr_score":0.06941105234243328,"valid_fraction":0.9900332225913622,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.4248358553276474,"r2_score":-1.3563303286586321,"spearmanr_score":0.17872306127281457,"valid_fraction":1,"rank":3}],"name":"MAPKAPK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mknk":{"meta":{"title":"Kinases Affinity Suite: MKNK","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mknk_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0260275004236032,"r2_score":0.004889022573936863,"spearmanr_score":0.4189226938174451,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.9567058248748292,"r2_score":0.14712163170991854,"spearmanr_score":0.39658116464474535,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9657357982923507,"r2_score":0.11422883946179507,"spearmanr_score":0.5373501873372023,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.080952586724633,"r2_score":-0.19898362823574334,"spearmanr_score":0.38075176296902,"valid_fraction":1,"rank":4}],"name":"MKNK","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-mknk2":{"meta":{"title":"Kinases Affinity Suite: MKNK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"mknk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.791889216506806,"r2_score":0.1623676852096958,"spearmanr_score":0.41356089595009243,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.803768886105538,"r2_score":0.2675120583023295,"spearmanr_score":0.47424526689229474,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.7562133019211015,"r2_score":0.24446823932467254,"spearmanr_score":0.5066688231565676,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.9621235729223387,"r2_score":-0.5024174334688394,"spearmanr_score":0.35509279156070095,"valid_fraction":1,"rank":4}],"name":"MKNK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-ntrk-1":{"meta":{"title":"Kinases Affinity Suite: NTRK-1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"ntrk-1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0054374287336951,"r2_score":-0.02598844719826543,"spearmanr_score":0.2959234933105105,"valid_fraction":0.9920544835414302,"rank":4},{"model":"Grok 4.5, high","mean_absolute_error":0.9601932559837344,"r2_score":0.10936388998992164,"spearmanr_score":0.3233592112646224,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9996395891121874,"r2_score":-0.05654341374693095,"spearmanr_score":0.34149560205067386,"valid_fraction":0.9875141884222475,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.9970793888794625,"r2_score":-0.04915314776520763,"spearmanr_score":0.4239056872927734,"valid_fraction":1,"rank":2}],"name":"NTRK-1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-pdgfr":{"meta":{"title":"Kinases Affinity Suite: PDGFR","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"pdgfr_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.3925971305260807,"r2_score":0.2965046575630238,"spearmanr_score":0.6048718543315393,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.40958429450949513,"r2_score":0.34033813122282763,"spearmanr_score":0.63606577284279,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.3825732339264031,"r2_score":0.27581578954499886,"spearmanr_score":0.6091301493404322,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.47680377718705874,"r2_score":0.06589408856047974,"spearmanr_score":0.5242554580078957,"valid_fraction":1,"rank":4}],"name":"PDGFR","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-pdpk1":{"meta":{"title":"Kinases Affinity Suite: PDPK1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"pdpk1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.6837535052720666,"r2_score":-0.3508270802331288,"spearmanr_score":0.25492477891716975,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":0.705749011590039,"r2_score":-0.05370053824547383,"spearmanr_score":0.20966582336246692,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":0.727155336955491,"r2_score":-0.4968211316006679,"spearmanr_score":0.26332020656482624,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.9772908794295864,"r2_score":-1.137599161890694,"spearmanr_score":0.296652839706803,"valid_fraction":1,"rank":4}],"name":"PDPK1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-pik3ca":{"meta":{"title":"Kinases Affinity Suite: PIK3CA","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"pik3ca_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0311345154710723,"r2_score":-0.21029245959472553,"spearmanr_score":0.31631460495898445,"valid_fraction":0.9962121212121212,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":1.089444919834734,"r2_score":-0.07821167851784305,"spearmanr_score":0.1672192722599887,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":1.3316927495073072,"r2_score":-1.314902384532732,"spearmanr_score":0.1178974862866655,"valid_fraction":0.9933712121212122,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.4227458231774026,"r2_score":-1.2452431720610755,"spearmanr_score":0.20102421411162416,"valid_fraction":1,"rank":4}],"name":"PIK3CA","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-pik3cg":{"meta":{"title":"Kinases Affinity Suite: PIK3CG","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"pik3cg_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.979582664380448,"r2_score":-0.42046441020598047,"spearmanr_score":0.16178104985809386,"valid_fraction":0.9981108312342569,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9167280925341573,"r2_score":-0.03995189324523385,"spearmanr_score":0.1597056779150397,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.168953269207386,"r2_score":-1.1282999272077099,"spearmanr_score":0.1064264679145447,"valid_fraction":0.9949622166246851,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":2.0029090615180314,"r2_score":-238.6131158068946,"spearmanr_score":0.15693072372632882,"valid_fraction":1,"rank":4}],"name":"PIK3CG","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-plk1":{"meta":{"title":"Kinases Affinity Suite: PLK1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"plk1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0666798690922372,"r2_score":-0.4908752303987969,"spearmanr_score":0.15415281724078073,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9716686800919289,"r2_score":-0.15898238715963764,"spearmanr_score":0.15561101187009022,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.2620588139430493,"r2_score":-0.9970655533021056,"spearmanr_score":0.3009637078378159,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":216.29583982210295,"r2_score":-10880479.664341109,"spearmanr_score":0.388148248139535,"valid_fraction":1,"rank":4}],"name":"PLK1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-plk2":{"meta":{"title":"Kinases Affinity Suite: PLK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"plk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.40891052010492573,"r2_score":0.01591183232549065,"spearmanr_score":0.5357116999373148,"valid_fraction":0.7608695652173914,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.4319018301705336,"r2_score":-0.08421060237299316,"spearmanr_score":0.25894160330790905,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.5073070855875518,"r2_score":-0.9889273003831336,"spearmanr_score":0.4458219018156762,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.3705369252163392,"r2_score":-0.2934609605567071,"spearmanr_score":0.4792822777328378,"valid_fraction":1,"rank":1}],"name":"PLK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-plk4":{"meta":{"title":"Kinases Affinity Suite: PLK4","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"plk4_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9618025629874523,"r2_score":-0.2781817973080629,"spearmanr_score":0.2259219792609931,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.8943996981363206,"r2_score":-0.01684275851061212,"spearmanr_score":0.3044932483857606,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9590483840956101,"r2_score":-0.31046005790888276,"spearmanr_score":0.3008847761752202,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.048035326537756,"r2_score":-0.8356409334451234,"spearmanr_score":0.3735535206849315,"valid_fraction":1,"rank":4}],"name":"PLK4","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-prkca":{"meta":{"title":"Kinases Affinity Suite: PRKCA","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"prkca_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9894364054401411,"r2_score":-0.28534700490980636,"spearmanr_score":0.14891577501208464,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9091902085777018,"r2_score":-0.034740308776193496,"spearmanr_score":0.08194778864189649,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.123014805841791,"r2_score":-0.7156622026680581,"spearmanr_score":0.23270156137509393,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":76.4631262056932,"r2_score":-443499.1115191216,"spearmanr_score":0.19588242495286395,"valid_fraction":1,"rank":4}],"name":"PRKCA","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-prkdc":{"meta":{"title":"Kinases Affinity Suite: PRKDC","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"prkdc_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.2219421659997958,"r2_score":-0.08661703948593535,"spearmanr_score":0.3472248249256268,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.1708168263909353,"r2_score":0.009343612491261033,"spearmanr_score":0.23449151276290583,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.3794152537549778,"r2_score":-0.5328007487996465,"spearmanr_score":0.2903033027109906,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.3780030060737516,"r2_score":-0.444253874477071,"spearmanr_score":0.39986944271615815,"valid_fraction":1,"rank":3}],"name":"PRKDC","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-ptk2":{"meta":{"title":"Kinases Affinity Suite: PTK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"ptk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.7619319507402524,"r2_score":-0.42634316718736676,"spearmanr_score":0.2286402791602066,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.6924791204870638,"r2_score":-0.05915756957674434,"spearmanr_score":0.13444634907585704,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.7941354181206987,"r2_score":-0.8205109687799503,"spearmanr_score":0.3007247339574083,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.9153869756348929,"r2_score":-1.067401909466291,"spearmanr_score":0.09106584026975062,"valid_fraction":1,"rank":4}],"name":"PTK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-ret":{"meta":{"title":"Kinases Affinity Suite: RET","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"ret_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.881754454229367,"r2_score":0.0011844110924988938,"spearmanr_score":0.22739832031645024,"valid_fraction":0.9878640776699029,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.8739210080451276,"r2_score":0.0904475163438111,"spearmanr_score":0.23373870131276717,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.0094471730260777,"r2_score":-7.465912998268982,"spearmanr_score":0.30259368083547966,"valid_fraction":0.9951456310679612,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":13.215317668442065,"r2_score":-80383.90051962451,"spearmanr_score":0.3950020258833929,"valid_fraction":1,"rank":4}],"name":"RET","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-rock2":{"meta":{"title":"Kinases Affinity Suite: ROCK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"rock2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.0809483710358483,"r2_score":-0.2835392790614546,"spearmanr_score":0.11126933335213654,"valid_fraction":0.9891518737672583,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.0221504183195083,"r2_score":-0.07686363638722771,"spearmanr_score":0.0901969693730843,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1324012391622733,"r2_score":-0.4135260745773335,"spearmanr_score":0.19820133488968905,"valid_fraction":0.9911242603550295,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.1876425145858513,"r2_score":-0.9147242585980206,"spearmanr_score":0.16156233813064733,"valid_fraction":1,"rank":4}],"name":"ROCK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-rsk2":{"meta":{"title":"Kinases Affinity Suite: RSK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"rsk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9155598684319831,"r2_score":-0.28385488163744044,"spearmanr_score":0.1169732715648579,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","mean_absolute_error":0.8422152343234952,"r2_score":0.016126100514744612,"spearmanr_score":0.2525392818774297,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.8761913300480593,"r2_score":-0.1879181894005133,"spearmanr_score":0.301903087910526,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.2547885624563209,"r2_score":-2.590827266676672,"spearmanr_score":0.06123942360000392,"valid_fraction":1,"rank":4}],"name":"RSK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-tek":{"meta":{"title":"Kinases Affinity Suite: TEK","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"tek_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.112948494249379,"r2_score":-0.8019307115356211,"spearmanr_score":0.12906679037864416,"valid_fraction":0.9938208032955715,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9441833477587898,"r2_score":-0.09403322471122033,"spearmanr_score":0.12581188822735462,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.3209662510243565,"r2_score":-1.7085229662391628,"spearmanr_score":0.11285120232853245,"valid_fraction":0.9938208032955715,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.2539587628556033,"r2_score":-1.2228802780813846,"spearmanr_score":0.22876854938223987,"valid_fraction":1,"rank":3}],"name":"TEK","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-tgfbr1":{"meta":{"title":"Kinases Affinity Suite: TGFBR1","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"tgfbr1_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.8965528707603402,"r2_score":0.19134442027516907,"spearmanr_score":0.6276220170442198,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","mean_absolute_error":1.0207143127893425,"r2_score":0.23410497178897438,"spearmanr_score":0.5070513656791279,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":1.039964120219197,"r2_score":-0.016404936684579763,"spearmanr_score":0.5593456857810511,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0357565828428987,"r2_score":-0.007793975562146338,"spearmanr_score":0.5548079907026143,"valid_fraction":1,"rank":3}],"name":"TGFBR1","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-tnik":{"meta":{"title":"Kinases Affinity Suite: TNIK","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"tnik_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.313263247681779,"r2_score":-0.3436265017068654,"spearmanr_score":0.22275346674625876,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","mean_absolute_error":1.1491299319037467,"r2_score":0.04975243086033154,"spearmanr_score":0.38043915419239627,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.1527664943310492,"r2_score":0.005401265252803378,"spearmanr_score":0.35242216011128885,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.2583025370539171,"r2_score":-0.16559883471091963,"spearmanr_score":0.3066438230382848,"valid_fraction":1,"rank":3}],"name":"TNIK","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-ttk":{"meta":{"title":"Kinases Affinity Suite: TTK","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"ttk_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9959498674845532,"r2_score":-0.08936905508075155,"spearmanr_score":0.2449724134238188,"valid_fraction":0.993933265925177,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.989346848727614,"r2_score":-0.014454955148258675,"spearmanr_score":0.17559347413697546,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9990596488458007,"r2_score":-0.11159042165240374,"spearmanr_score":0.29823097672729443,"valid_fraction":0.9949443882709808,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.247289899724865,"r2_score":-21.24626056777858,"spearmanr_score":0.2968024493080976,"valid_fraction":1,"rank":4}],"name":"TTK","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-tyk2":{"meta":{"title":"Kinases Affinity Suite: TYK2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"tyk2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.9377148417213855,"r2_score":-0.18381277142921637,"spearmanr_score":0.10722599164637316,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.9062613012631041,"r2_score":-0.09501210700532514,"spearmanr_score":0.07227974098859868,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":0.9656448885782231,"r2_score":-0.22107565599207013,"spearmanr_score":0.026375271630795304,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.0513876585161903,"r2_score":-0.6446855039882804,"spearmanr_score":0.11336006530822793,"valid_fraction":1,"rank":4}],"name":"TYK2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-vegfr2":{"meta":{"title":"Kinases Affinity Suite: VEGFR2","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"vegfr2_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":1.3721238729164729,"r2_score":-0.6347061377341903,"spearmanr_score":0.14227554843209286,"valid_fraction":0.9856076759061834,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":1.1984215097505446,"r2_score":-0.06267442055099126,"spearmanr_score":0.17771340651254874,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","mean_absolute_error":1.4810699323081773,"r2_score":-0.9518212580820911,"spearmanr_score":0.3030485564832422,"valid_fraction":0.9845415778251599,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":1.4867158422065812,"r2_score":-1.664671001023089,"spearmanr_score":0.24422465359469767,"valid_fraction":1,"rank":4}],"name":"VEGFR2","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"med-kinases-affinity-suite-zap70":{"meta":{"title":"Kinases Affinity Suite: ZAP70","category":"Medicinal Chemistry > Kinases Affinity Suite","description":"zap70_reg_pred_single","metric_types":[],"paper_link":"#","source_file":"kinases_ci-030826.json","original_category":"Medicinal Chemistry"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Claude Opus 4.8, high","mean_absolute_error":0.6957739964633847,"r2_score":-0.4208854111069862,"spearmanr_score":0.07949933481406744,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":0.7323107040242103,"r2_score":-0.23913146377799688,"spearmanr_score":0.0517079410306125,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","mean_absolute_error":0.6930287056702444,"r2_score":-0.479435079929893,"spearmanr_score":0.09282791174090267,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":0.8242220464778697,"r2_score":-0.7478221223802792,"spearmanr_score":0.06907528152640005,"valid_fraction":1,"rank":4}],"name":"ZAP70","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"spearmanr_score","min":-1,"max":1}},"che-retrosynthesis-suite-single-step-retrosynthesis-ursa-expert-2026-dataset":{"meta":{"title":"Single-step retrosynthesis, Top-1 task: Single-step retrosynthesis, URSA-expert-2026 dataset","category":"Chemical Synthesis > Single-step retrosynthesis, Top-1 task","description":"Evaluates single-step retrosynthesis task performance on URSA-expert-2026 dataset across various ChemCensor metrics.","metric_types":[],"paper_link":"#","source_file":"retrosynthesis_ci.json","original_category":"Synthesis"},"radar_keys":["apt_max_cc","apt_top_3","apt_top_5","apt_top_10"],"leaderboard_metric":"apt_max_cc","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","apt_max_cc":1.54,"apt_top_3":1.0367,"apt_top_5":0.698,"apt_top_10":0.353,"rank":3},{"model":"GPT 5.5, medium","model_type":"llm","apt_max_cc":1.54,"apt_top_3":0.9867,"apt_top_5":0.648,"apt_top_10":0.327,"rank":3},{"model":"Gemini 3.1 Pro, high","model_type":"llm","apt_max_cc":1.5,"apt_top_3":0.8167,"apt_top_5":0.506,"apt_top_10":0.253,"rank":5},{"model":"Grok 4.5, high","model_type":"llm","apt_max_cc":1.44,"apt_top_3":0.86,"apt_top_5":0.558,"apt_top_10":0.28,"rank":6},{"model":"Claude Opus 4.8, high","model_type":"llm","apt_max_cc":1.7,"apt_top_3":1.1133,"apt_top_5":0.72,"apt_top_10":0.362,"rank":2},{"model":"RetroChimera","model_type":"specialist","apt_max_cc":1.98,"apt_top_3":1.78,"apt_top_5":1.54,"apt_top_10":1.11,"rank":1}],"name":"Single-step retrosynthesis, URSA-expert-2026 dataset","metric_info":{"apt_max_cc":{"label":"APT, max CC","shortLabel":"APT-Max","direction":"higher_better","description":"Per-target maximum ChemCensor score averaged over targets.","units":"absolute"},"apt_top_3":{"label":"APT, top-3","shortLabel":"APT-Top-3","direction":"higher_better","description":"Per-target average ChemCensor score over top-3 unique predictions.","units":"absolute"},"apt_top_5":{"label":"APT, top-5","shortLabel":"APT-Top-5","direction":"higher_better","description":"Per-target average ChemCensor score over top-5 unique predictions.","units":"absolute"},"apt_top_10":{"label":"APT, top-10","shortLabel":"APT-Top-10","direction":"higher_better","description":"Per-target average ChemCensor score over top-10 unique predictions.","units":"absolute"}},"insilico_score":{"metric":"apt_top_10","min":0,"max":5}},"che-retrosynthesis-suite-single-step-retrosynthesis-uspto-50k-test-sample-dataset":{"meta":{"title":"Single-step retrosynthesis, Top-1 task: Single-step retrosynthesis, USPTO-50k-test sample dataset","category":"Chemical Synthesis > Single-step retrosynthesis, Top-1 task","description":"Evaluates single-step retrosynthesis task performance on representative 10% sample from USPTO-50k-test dataset across various ChemCensor metrics.","metric_types":[],"paper_link":"#","source_file":"retrosynthesis_ci.json","original_category":"Synthesis"},"radar_keys":["apt_max_cc","apt_top_3","apt_top_5","apt_top_10"],"leaderboard_metric":"apt_max_cc","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","apt_max_cc":3.9698,"apt_top_3":2.0812,"apt_top_5":1.3219,"apt_top_10":0.6658,"rank":2},{"model":"GPT 5.5, medium","model_type":"llm","apt_max_cc":3.9235,"apt_top_3":1.9611,"apt_top_5":1.2334,"apt_top_10":0.6189,"rank":3},{"model":"Gemini 3.1 Pro, high","model_type":"llm","apt_max_cc":3.9054,"apt_top_3":1.6137,"apt_top_5":0.9694,"apt_top_10":0.4847,"rank":4},{"model":"Grok 4.5, high","model_type":"llm","apt_max_cc":3.8129,"apt_top_3":1.782,"apt_top_5":1.1022,"apt_top_10":0.5535,"rank":5},{"model":"Claude Opus 4.8, high","model_type":"llm","apt_max_cc":3.7389,"apt_top_3":1.7834,"apt_top_5":1.0988,"apt_top_10":0.5508,"rank":6},{"model":"RetroChimera","model_type":"specialist","apt_max_cc":4.8672,"apt_top_3":3.3239,"apt_top_5":2.6209,"apt_top_10":1.6529,"rank":1}],"name":"Single-step retrosynthesis, USPTO-50k-test sample dataset","metric_info":{"apt_max_cc":{"label":"APT, max CC","shortLabel":"APT-Max","direction":"higher_better","description":"Per-target maximum ChemCensor score averaged over targets.","units":"absolute"},"apt_top_3":{"label":"APT, top-3","shortLabel":"APT-Top-3","direction":"higher_better","description":"Per-target average ChemCensor score over top-3 unique predictions.","units":"absolute"},"apt_top_5":{"label":"APT, top-5","shortLabel":"APT-Top-5","direction":"higher_better","description":"Per-target average ChemCensor score over top-5 unique predictions.","units":"absolute"},"apt_top_10":{"label":"APT, top-10","shortLabel":"APT-Top-10","direction":"higher_better","description":"Per-target average ChemCensor score over top-10 unique predictions.","units":"absolute"}},"insilico_score":{"metric":"apt_top_10","min":0,"max":5}},"syn-retrosynthesis-suite-single-step-retrosynthesis-top-15-task-ursa-expert-2026-dataset":{"meta":{"title":"Single-step retrosynthesis, Top-15: Single-step retrosynthesis, Top-15 task, URSA-expert-2026 dataset","category":"Chemical Synthesis > Single-step retrosynthesis, Top-15","description":"Top-k single-step retrosynthesis on the URSA-expert-2026 test set. The model returns a ranked list of up to k=15 predicted reactant sets per target; the prompt is repeated over 3 independent trials and ChemCensor metrics are computed per trial and averaged.","metric_types":[],"paper_link":"#","source_file":"retrosynthesis_topk_ci.json","original_category":"Synthesis"},"radar_keys":["apt_max_cc","apt_top_3","apt_top_5","apt_top_10"],"leaderboard_metric":"apt_max_cc","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","apt_max_cc":2.0267,"apt_top_3":1.7622,"apt_top_5":1.5213,"apt_top_10":1.1197,"rank":2},{"model":"GPT 5.5, medium","model_type":"llm","apt_max_cc":1.94,"apt_top_3":1.68,"apt_top_5":1.46,"apt_top_10":1.05,"rank":6},{"model":"Gemini 3.1 Pro, high","model_type":"llm","apt_max_cc":1.91,"apt_top_3":1.69,"apt_top_5":1.46,"apt_top_10":1.08,"rank":8},{"model":"GPT 5.6 Terra, high","model_type":"llm","apt_max_cc":1.8633,"apt_top_3":1.6256,"apt_top_5":1.3947,"apt_top_10":0.9937,"rank":10},{"model":"Grok 4.5, high","model_type":"llm","apt_max_cc":1.92,"apt_top_3":1.67,"apt_top_5":1.43,"apt_top_10":1.03,"rank":7},{"model":"Claude Opus 4.8, high","model_type":"llm","apt_max_cc":1.89,"apt_top_3":1.62,"apt_top_5":1.34,"apt_top_10":0.85,"rank":9},{"model":"Kimi K3, max","model_type":"llm","apt_max_cc":1.6064,"apt_top_3":1.2626,"apt_top_5":1.0112,"apt_top_10":0.6168,"rank":13},{"model":"GPT 5.6 Luna, high","model_type":"llm","apt_max_cc":1.3867,"apt_top_3":1.2156,"apt_top_5":1.0667,"apt_top_10":0.7337,"rank":14},{"model":"GPT 5.6 Terra, medium","model_type":"llm","apt_max_cc":1.6486,"apt_top_3":1.3822,"apt_top_5":1.1491,"apt_top_10":0.717,"rank":12},{"model":"GPT 5.6 Luna, medium","model_type":"llm","apt_max_cc":1.0471,"apt_top_3":0.9349,"apt_top_5":0.8418,"apt_top_10":0.6044,"rank":15},{"model":"RetroChimera","model_type":"specialist","apt_max_cc":1.98,"apt_top_3":1.78,"apt_top_5":1.54,"apt_top_10":1.11,"rank":5},{"model":"LFM2-MMAI-Chem-SSRS","model_type":"specialist","apt_max_cc":2.09,"apt_top_3":1.85,"apt_top_5":1.59,"apt_top_10":1.11,"rank":1},{"model":"GPT 5.6 Sol, high","apt_max_cc":1.9833333333333332,"apt_top_3":1.7544444444444436,"apt_top_5":1.518,"apt_top_10":1.1303333333333334,"rank":4},{"model":"Claude Opus 5, high","apt_max_cc":1.9896556560768637,"apt_top_3":1.7401716557184155,"apt_top_5":1.5073588026275804,"apt_top_10":1.1135169016060031,"rank":3},{"model":"Qwen 3.8 Max","apt_max_cc":1.7236769759450172,"apt_top_3":1.463081328751432,"apt_top_5":1.248769759450172,"apt_top_10":0.8471443298969072,"rank":11}],"name":"Single-step retrosynthesis, Top-15 task, URSA-expert-2026 dataset","metric_info":{"apt_max_cc":{"label":"APT, max CC","shortLabel":"APT-Max","direction":"higher_better","description":"Per-target maximum ChemCensor score averaged over targets.","units":"absolute"},"apt_top_3":{"label":"APT, top-3","shortLabel":"APT-Top-3","direction":"higher_better","description":"Per-target average ChemCensor score over top-3 unique predictions.","units":"absolute"},"apt_top_5":{"label":"APT, top-5","shortLabel":"APT-Top-5","direction":"higher_better","description":"Per-target average ChemCensor score over top-5 unique predictions.","units":"absolute"},"apt_top_10":{"label":"APT, top-10","shortLabel":"APT-Top-10","direction":"higher_better","description":"Per-target average ChemCensor score over top-10 unique predictions.","units":"absolute"}},"insilico_score":{"metric":"apt_top_10","min":0,"max":5}},"syn-retrosynthesis-suite-single-step-retrosynthesis-top-15-task-uspto-50k-test-sample-dataset":{"meta":{"title":"Single-step retrosynthesis, Top-15: Single-step retrosynthesis, Top-15 task, USPTO-50k-test sample dataset","category":"Chemical Synthesis > Single-step retrosynthesis, Top-15","description":"Top-k single-step retrosynthesis on the USPTO-50k-test sample. The model returns a ranked list of up to k=15 predicted reactant sets per target; the prompt is repeated over 3 independent trials and ChemCensor metrics are computed per trial and averaged.","metric_types":[],"paper_link":"#","source_file":"retrosynthesis_topk_ci.json","original_category":"Synthesis"},"radar_keys":["apt_max_cc","apt_top_3","apt_top_5","apt_top_10"],"leaderboard_metric":"apt_max_cc","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","apt_max_cc":4.5828,"apt_top_3":3.1552,"apt_top_5":2.5459,"apt_top_10":1.7066,"rank":4},{"model":"GPT 5.5, medium","model_type":"llm","apt_max_cc":4.5,"apt_top_3":3.07,"apt_top_5":2.45,"apt_top_10":1.63,"rank":5},{"model":"Gemini 3.1 Pro, high","model_type":"llm","apt_max_cc":4.32,"apt_top_3":2.92,"apt_top_5":2.34,"apt_top_10":1.59,"rank":9},{"model":"GPT 5.6 Terra, high","model_type":"llm","apt_max_cc":4.323,"apt_top_3":3.0287,"apt_top_5":2.4294,"apt_top_10":1.5842,"rank":8},{"model":"Grok 4.5, high","model_type":"llm","apt_max_cc":4.47,"apt_top_3":3.03,"apt_top_5":2.43,"apt_top_10":1.62,"rank":6},{"model":"Claude Opus 4.8, high","model_type":"llm","apt_max_cc":4.35,"apt_top_3":2.95,"apt_top_5":2.3,"apt_top_10":1.44,"rank":7},{"model":"Kimi K3, max","model_type":"llm","apt_max_cc":3.9219,"apt_top_3":2.622,"apt_top_5":2.0281,"apt_top_10":1.24,"rank":11},{"model":"GPT 5.6 Luna, high","model_type":"llm","apt_max_cc":3.8208,"apt_top_3":2.9753,"apt_top_5":2.5128,"apt_top_10":1.6968,"rank":12},{"model":"GPT 5.6 Terra, medium","model_type":"llm","apt_max_cc":3.716,"apt_top_3":2.594,"apt_top_5":2.0475,"apt_top_10":1.2701,"rank":13},{"model":"GPT 5.6 Luna, medium","model_type":"llm","apt_max_cc":3.1593,"apt_top_3":2.5707,"apt_top_5":2.229,"apt_top_10":1.562,"rank":15},{"model":"RetroChimera","model_type":"specialist","apt_max_cc":4.8672,"apt_top_3":3.3239,"apt_top_5":2.6209,"apt_top_10":1.6529,"rank":1},{"model":"LFM2-MMAI-Chem-SSRS","model_type":"specialist","apt_max_cc":3.66,"apt_top_3":2.54,"apt_top_5":2.06,"apt_top_10":1.42,"rank":14},{"model":"GPT 5.6 Sol, high","apt_max_cc":4.608987256874581,"apt_top_3":3.154482450257094,"apt_top_5":2.540845070422535,"apt_top_10":1.7102615694164995,"rank":3},{"model":"Claude Opus 5, high","apt_max_cc":4.616008198151055,"apt_top_3":3.2240842490842474,"apt_top_5":2.583599337170766,"apt_top_10":1.7280651491365777,"rank":2},{"model":"Qwen 3.8 Max","apt_max_cc":4.200286123032904,"apt_top_3":2.894611349546972,"apt_top_5":2.292846924177397,"apt_top_10":1.4749642346208869,"rank":10}],"name":"Single-step retrosynthesis, Top-15 task, USPTO-50k-test sample dataset","metric_info":{"apt_max_cc":{"label":"APT, max CC","shortLabel":"APT-Max","direction":"higher_better","description":"Per-target maximum ChemCensor score averaged over targets.","units":"absolute"},"apt_top_3":{"label":"APT, top-3","shortLabel":"APT-Top-3","direction":"higher_better","description":"Per-target average ChemCensor score over top-3 unique predictions.","units":"absolute"},"apt_top_5":{"label":"APT, top-5","shortLabel":"APT-Top-5","direction":"higher_better","description":"Per-target average ChemCensor score over top-5 unique predictions.","units":"absolute"},"apt_top_10":{"label":"APT, top-10","shortLabel":"APT-Top-10","direction":"higher_better","description":"Per-target average ChemCensor score over top-10 unique predictions.","units":"absolute"}},"insilico_score":{"metric":"apt_top_10","min":0,"max":5}},"syn-msrs-multi-step-retrosynthesis-ursa-expert-2026-test-set":{"meta":{"title":"Multi-step retrosynthesis: Multi-step retrosynthesis, URSA-expert-2026 test set","category":"Chemical Synthesis > Multi-step retrosynthesis","description":"Evaluates multi-step retrosynthesis performance on the URSA-expert-2026 test set across Solv-N and average per-route ChemCensor metrics.","metric_types":[],"paper_link":"#","source_file":"msrs_ci.json","original_category":"Synthesis"},"radar_keys":["solv_0","solv_1","solv_2","av_pr_cc"],"leaderboard_metric":"solv_0","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","solv_0":72,"solv_1":42,"solv_2":17,"av_pr_cc":1.7018,"rank":4},{"model":"GPT 5.5, medium","model_type":"llm","solv_0":63,"solv_1":44,"solv_2":21,"av_pr_cc":1.62,"rank":8},{"model":"Gemini 3.1 Pro, high","model_type":"llm","solv_0":54,"solv_1":30,"solv_2":13,"av_pr_cc":1.46,"rank":13},{"model":"GPT 5.6 Terra, high","model_type":"llm","solv_0":73,"solv_1":39,"solv_2":19,"av_pr_cc":1.5877,"rank":3},{"model":"Grok 4.5, high","model_type":"llm","solv_0":65,"solv_1":34,"solv_2":15,"av_pr_cc":1.3971,"rank":6},{"model":"Claude Opus 4.8, high","model_type":"llm","solv_0":55,"solv_1":19,"solv_2":10,"av_pr_cc":1.3,"rank":11},{"model":"Kimi K3, max","model_type":"llm","solv_0":55,"solv_1":19,"solv_2":8,"av_pr_cc":1.2217,"rank":11},{"model":"GPT 5.6 Luna, high","model_type":"llm","solv_0":59,"solv_1":33,"solv_2":14,"av_pr_cc":1.4601,"rank":9},{"model":"GPT 5.6 Terra, medium","model_type":"llm","solv_0":59,"solv_1":15,"solv_2":8,"av_pr_cc":1.1224,"rank":9},{"model":"GPT 5.6 Luna, medium","model_type":"llm","solv_0":39,"solv_1":8,"solv_2":4,"av_pr_cc":1.15,"rank":15},{"model":"RetroChimera-MCTS","model_type":"specialist","solv_0":99,"solv_1":78,"solv_2":32,"av_pr_cc":1.5303,"rank":1},{"model":"Aizynthfinder-MCTS","model_type":"specialist","solv_0":65,"solv_1":58,"solv_2":29,"av_pr_cc":1.61,"rank":6},{"model":"GPT 5.6 Sol, high","solv_0":71,"solv_1":48,"solv_2":26,"av_pr_cc":1.6838235294117647,"rank":5},{"model":"Claude Opus 5, high","solv_0":52,"solv_1":39,"solv_2":19,"av_pr_cc":1.6681818181818182,"rank":14},{"model":"Qwen 3.8 Max","solv_0":81,"solv_1":44,"solv_2":15,"av_pr_cc":1.6904761904761905,"rank":2}],"name":"Multi-step retrosynthesis, URSA-expert-2026 test set","metric_info":{"solv_0":{"label":"Solv-0","shortLabel":"Solv-0","direction":"higher_better","description":"Fraction of targets solved at Solv-0 level, reported as a percentage. The route contains valid SMILES and terminates in compounds available in the starting-material stock. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"},"solv_1":{"label":"Solv-1","shortLabel":"Solv-1","direction":"higher_better","description":"Fraction of targets solved at Solv-1 level, reported as a percentage. Each reaction must be valid and supported by reaction-center precedent. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"},"solv_2":{"label":"Solv-2","shortLabel":"Solv-2","direction":"higher_better","description":"Fraction of targets solved at Solv-2 level, reported as a percentage. Each reaction must be additionally supported by functional-group compatibility. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"},"av_pr_cc":{"label":"Av. PR CC","shortLabel":"Av. PR CC","direction":"higher_better","description":"Average per-reaction ChemCensor score from the best per target route. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"}},"insilico_score":{"metric":"solv_2","min":0,"max":100}},"syn-msrs-multi-step-retrosynthesis-ursa-drugsclinicals-2026-test-set":{"meta":{"title":"Multi-step retrosynthesis: Multi-step retrosynthesis, URSA-drugs&clinicals-2026 test set","category":"Chemical Synthesis > Multi-step retrosynthesis","description":"Evaluates multi-step retrosynthesis performance on the URSA-drugs&clinicals-2026 test set across Solv-N and average per-route ChemCensor metrics.","metric_types":[],"paper_link":"#","source_file":"msrs_ci.json","original_category":"Synthesis"},"radar_keys":["solv_0","solv_1","solv_2","av_pr_cc"],"leaderboard_metric":"solv_0","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","solv_0":83,"solv_1":65,"solv_2":41,"av_pr_cc":2.4582,"rank":2},{"model":"GPT 5.5, medium","model_type":"llm","solv_0":75,"solv_1":54,"solv_2":35,"av_pr_cc":2.43,"rank":9},{"model":"Gemini 3.1 Pro, high","model_type":"llm","solv_0":70,"solv_1":48,"solv_2":38,"av_pr_cc":2.42,"rank":12},{"model":"GPT 5.6 Terra, high","model_type":"llm","solv_0":78,"solv_1":49,"solv_2":32,"av_pr_cc":2.2983,"rank":5},{"model":"Grok 4.5, high","model_type":"llm","solv_0":77,"solv_1":48,"solv_2":31,"av_pr_cc":2.2072,"rank":7},{"model":"Claude Opus 4.8, high","model_type":"llm","solv_0":75,"solv_1":37,"solv_2":27,"av_pr_cc":1.99,"rank":9},{"model":"Kimi K3, max","model_type":"llm","solv_0":76,"solv_1":37,"solv_2":28,"av_pr_cc":2.0036,"rank":8},{"model":"GPT 5.6 Luna, high","model_type":"llm","solv_0":73,"solv_1":48,"solv_2":30,"av_pr_cc":2.272,"rank":11},{"model":"GPT 5.6 Terra, medium","model_type":"llm","solv_0":61,"solv_1":28,"solv_2":17,"av_pr_cc":2.0289,"rank":14},{"model":"GPT 5.6 Luna, medium","model_type":"llm","solv_0":55,"solv_1":30,"solv_2":20,"av_pr_cc":2.0335,"rank":15},{"model":"RetroChimera-MCTS","model_type":"specialist","solv_0":100,"solv_1":83,"solv_2":60,"av_pr_cc":2.2481,"rank":1},{"model":"Aizynthfinder-MCTS","model_type":"specialist","solv_0":80,"solv_1":71,"solv_2":54,"av_pr_cc":2.29,"rank":3},{"model":"GPT 5.6 Sol, high","solv_0":78,"solv_1":57.99999999999999,"solv_2":38,"av_pr_cc":2.4160206718346253,"rank":5},{"model":"Claude Opus 5, high","solv_0":70,"solv_1":57.99999999999999,"solv_2":39,"av_pr_cc":2.5753424657534247,"rank":12},{"model":"Qwen 3.8 Max","solv_0":79,"solv_1":52,"solv_2":30,"av_pr_cc":2.391011235955056,"rank":4}],"name":"Multi-step retrosynthesis, URSA-drugs&clinicals-2026 test set","metric_info":{"solv_0":{"label":"Solv-0","shortLabel":"Solv-0","direction":"higher_better","description":"Fraction of targets solved at Solv-0 level, reported as a percentage. The route contains valid SMILES and terminates in compounds available in the starting-material stock. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"},"solv_1":{"label":"Solv-1","shortLabel":"Solv-1","direction":"higher_better","description":"Fraction of targets solved at Solv-1 level, reported as a percentage. Each reaction must be valid and supported by reaction-center precedent. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"},"solv_2":{"label":"Solv-2","shortLabel":"Solv-2","direction":"higher_better","description":"Fraction of targets solved at Solv-2 level, reported as a percentage. Each reaction must be additionally supported by functional-group compatibility. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"},"av_pr_cc":{"label":"Av. PR CC","shortLabel":"Av. PR CC","direction":"higher_better","description":"Average per-reaction ChemCensor score from the best per target route. For details read our URSA preprint: https://arxiv.org/pdf/2607.04688.","units":"absolute"}},"insilico_score":{"metric":"solv_2","min":0,"max":100}},"che-reaction-plausibility-reaction-plausibility":{"meta":{"title":"Reaction Plausibility: Reaction Plausibility","category":"Chemical Synthesis > Reaction Plausibility","description":"Evaluates binary reaction plausibility classification on a balanced set of 1,000 model-predicted reactions (500 plausible, 500 implausible) from multiple LLMs, with expert labels confirmed in SciFinder.","metric_types":[],"paper_link":"#","source_file":"reaction_plausibility_ci.json","original_category":"Synthesis"},"radar_keys":["mcc_score","binary_f1_score","accuracy_score","validity"],"leaderboard_metric":"mcc_score","results":[{"model":"GPT 5.6 Sol, medium","model_type":"llm","mcc_score":0.3755,"binary_f1_score":0.7282,"accuracy_score":0.6305,"validity":0.998,"rank":9},{"model":"GPT 5.5, medium","model_type":"llm","mcc_score":0.5688,"binary_f1_score":0.801,"accuracy_score":0.764,"validity":1,"rank":3},{"model":"Gemini 3.1 Pro, high","model_type":"llm","mcc_score":0.669,"binary_f1_score":0.8434,"accuracy_score":0.828,"validity":1,"rank":1},{"model":"GPT 5.6 Terra, high","model_type":"llm","mcc_score":0.3162,"binary_f1_score":0.712,"accuracy_score":0.602,"validity":1,"rank":10},{"model":"Grok 4.5, high","model_type":"llm","mcc_score":0.6157,"binary_f1_score":0.8206,"accuracy_score":0.795,"validity":1,"rank":2},{"model":"Claude Opus 4.8, high","model_type":"llm","mcc_score":0.5236,"binary_f1_score":0.785,"accuracy_score":0.7333,"validity":1,"rank":4},{"model":"Kimi K3, max","model_type":"llm","mcc_score":0.3899,"binary_f1_score":0.7508,"accuracy_score":0.6536,"validity":0.941,"rank":7},{"model":"GPT 5.6 Luna, high","model_type":"llm","mcc_score":0.3133,"binary_f1_score":0.71,"accuracy_score":0.594,"validity":1,"rank":11},{"model":"GPT 5.6 Terra, medium","model_type":"llm","mcc_score":0.2897,"binary_f1_score":0.7055,"accuracy_score":0.591,"validity":1,"rank":13},{"model":"GPT 5.6 Luna, medium","model_type":"llm","mcc_score":0.31,"binary_f1_score":0.7106,"accuracy_score":0.6,"validity":1,"rank":12},{"model":"GPT 5.6 Sol, high","mcc_score":0.3844144171822536,"binary_f1_score":0.7303453343130052,"accuracy_score":0.6318956870611836,"validity":0.998,"rank":8},{"model":"Claude Opus 5, high","mcc_score":0.5094936865967384,"binary_f1_score":0.77663772691397,"accuracy_score":0.7138523761375126,"validity":0.989,"rank":5},{"model":"Qwen 3.8 Max","mcc_score":0.41291976927471014,"binary_f1_score":0.7420333839150227,"accuracy_score":0.6576032225579054,"validity":0.993,"rank":6}],"name":"Reaction Plausibility","metric_info":{"mcc_score":{"label":"Matthews correlation coefficient","shortLabel":"MCC","direction":"higher_better"},"binary_f1_score":{"label":"Binary F1 score","shortLabel":"BinF1","direction":"higher_better"},"accuracy_score":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better"},"validity":{"label":"Validity","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"accuracy_score","min":0,"max":1}},"cli-clinbench-quarterly-all-phases-2025q1":{"meta":{"title":"ClinBench Quarterly: All Phases - 2025Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q1. N=356 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8623595505617978,"balanced_accuracy":0.8623298272213272,"precision":0.8563218390804598,"recall":0.861271676300578,"f1":0.8587896253602305,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8087774294670846,"balanced_accuracy":0.8089668999135152,"precision":0.793939393939394,"recall":0.8291139240506329,"f1":0.8111455108359134,"valid_fraction":0.8960674157303371,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8239,"accuracy":0.8249,"f1":0.8144,"precision":0.8395,"recall":0.7907,"n_samples":356,"tp":136,"tn":156,"fp":26,"fn":36,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","accuracy":0.8061797752808989,"balanced_accuracy":0.8045263590132348,"precision":0.8376623376623377,"recall":0.7456647398843931,"f1":0.7889908256880734,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8351653558229888,"accuracy":0.8342696629213483,"f1":0.8356545961002786,"precision":0.8064516129032258,"recall":0.8670520231213873,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","balanced_accuracy":0.8719405594405594,"accuracy":0.8724637681159421,"f1":0.8666666666666667,"precision":0.8881987577639752,"recall":0.8461538461538461,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.9691011235955056,"rank":1}],"name":"All Phases - 2025Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-all-phases-2025q2":{"meta":{"title":"ClinBench Quarterly: All Phases - 2025Q2","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q2. N=454 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8436123348017621,"balanced_accuracy":0.8406832720838437,"precision":0.8842105263157894,"recall":0.7741935483870968,"f1":0.8255528255528255,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Claude Opus 4.8, high","accuracy":0.69,"balanced_accuracy":0.6933092948717949,"precision":0.6478260869565218,"recall":0.7760416666666666,"f1":0.7061611374407583,"valid_fraction":0.8810572687224669,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8197,"accuracy":0.8234,"f1":0.8,"precision":0.8696,"recall":0.7407,"n_samples":454,"tp":160,"tn":213,"fp":24,"fn":56,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy":0.8281938325991189,"balanced_accuracy":0.8231931400571662,"precision":0.9112426035502958,"recall":0.7096774193548387,"f1":0.7979274611398963,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.7369771918567345,"accuracy":0.7312775330396476,"f1":0.7550200803212851,"precision":0.6690391459074733,"recall":0.8663594470046083,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy":0.8134467833956325,"accuracy":0.8222222222222222,"f1":0.7857142857142857,"precision":0.868421052631579,"recall":0.717391304347826,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8920704845814978,"rank":4}],"name":"All Phases - 2025Q2","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-all-phases-2025q3":{"meta":{"title":"ClinBench Quarterly: All Phases - 2025Q3","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q3. N=297 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8417508417508418,"balanced_accuracy":0.827890534821085,"precision":0.8055555555555556,"recall":0.7699115044247787,"f1":0.7873303167420814,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8141263940520446,"balanced_accuracy":0.8065359477124183,"precision":0.7333333333333333,"recall":0.7777777777777778,"f1":0.7549019607843137,"valid_fraction":0.9057239057239057,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8036,"accuracy":0.8209,"f1":0.7558,"precision":0.781,"recall":0.7321,"n_samples":297,"tp":82,"tn":161,"fp":23,"fn":30,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.835016835016835,"balanced_accuracy":0.8105040400153904,"precision":0.8333333333333334,"recall":0.7079646017699115,"f1":0.7655502392344498,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8231531358214699,"accuracy":0.8316498316498316,"f1":0.7807017543859649,"precision":0.7739130434782608,"recall":0.7876106194690266,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","balanced_accuracy":0.8711227890284542,"accuracy":0.8884615384615384,"f1":0.8397790055248618,"precision":0.8735632183908046,"recall":0.8085106382978723,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8754208754208754,"rank":1}],"name":"All Phases - 2025Q3","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-all-phases-2025q4":{"meta":{"title":"ClinBench Quarterly: All Phases - 2025Q4","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (All Phases) for trials with results submitted in 2025Q4. N=293 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7986348122866894,"balanced_accuracy":0.7984382284382284,"precision":0.8013245033112583,"recall":0.8066666666666666,"f1":0.8039867109634552,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Claude Opus 4.8, high","accuracy":0.7953667953667953,"balanced_accuracy":0.7989410075385903,"precision":0.8559322033898306,"recall":0.7372262773722628,"f1":0.792156862745098,"valid_fraction":0.8839590443686007,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7966,"accuracy":0.7952,"f1":0.7872,"precision":0.8409,"recall":0.74,"n_samples":293,"tp":111,"tn":122,"fp":21,"fn":39,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.7986348122866894,"balanced_accuracy":0.8008857808857809,"precision":0.8760330578512396,"recall":0.7066666666666667,"f1":0.7822878228782287,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.808927738927739,"accuracy":0.8088737201365188,"f1":0.8120805369127517,"precision":0.8175675675675675,"recall":0.8066666666666666,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","balanced_accuracy":0.8591544715447155,"accuracy":0.8588709677419355,"f1":0.8547717842323651,"precision":0.8879310344827587,"recall":0.824,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8464163822525598,"rank":1}],"name":"All Phases - 2025Q4","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-all-phases-2026q1":{"meta":{"title":"ClinBench Quarterly: All Phases - 2026Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (All Phases) for trials with results submitted in 2026Q1. N=87 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8620689655172413,"balanced_accuracy":0.882996632996633,"precision":0.9772727272727273,"recall":0.7962962962962963,"f1":0.8775510204081632,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Claude Opus 4.8, high","accuracy":0.875,"balanced_accuracy":0.8796484110885734,"precision":0.9361702127659575,"recall":0.8627450980392157,"f1":0.8979591836734694,"valid_fraction":0.9195402298850575,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.782,"accuracy":0.7586,"f1":0.7789,"precision":0.9024,"recall":0.6852,"n_samples":87,"tp":37,"tn":29,"fp":4,"fn":17,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.8045977011494253,"balanced_accuracy":0.8249158249158248,"precision":0.9302325581395349,"recall":0.7407407407407407,"f1":0.8247422680412371,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8409090909090909,"accuracy":0.8390804597701149,"f1":0.8653846153846154,"precision":0.9,"recall":0.8333333333333334,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.8464240102171137,"accuracy":0.8313253012048193,"f1":0.86,"precision":0.9347826086956522,"recall":0.7962962962962963,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.9540229885057471,"rank":3}],"name":"All Phases - 2026Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-1-2025q1":{"meta":{"title":"ClinBench Quarterly: Phase 1 - 2025Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q1. N=56 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.875,"balanced_accuracy":0.8583333333333334,"precision":0.8421052631578947,"recall":0.8,"f1":0.8205128205128205,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.7090909090909091,"balanced_accuracy":0.7032163742690059,"precision":0.5652173913043478,"recall":0.6842105263157895,"f1":0.6190476190476191,"valid_fraction":0.9821428571428571,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7667,"accuracy":0.7857,"f1":0.7,"precision":0.7,"recall":0.7,"n_samples":56,"tp":14,"tn":30,"fp":6,"fn":6,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","accuracy":0.7321428571428571,"balanced_accuracy":0.6916666666666667,"precision":0.6470588235294118,"recall":0.55,"f1":0.5945945945945946,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8694444444444445,"accuracy":0.875,"f1":0.8292682926829268,"precision":0.8095238095238095,"recall":0.85,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy":0.8420138888888888,"accuracy":0.86,"f1":0.8,"precision":0.8235294117647058,"recall":0.7777777777777778,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8928571428571429,"rank":3}],"name":"Phase 1 - 2025Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-1-2025q2":{"meta":{"title":"ClinBench Quarterly: Phase 1 - 2025Q2","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q2. N=127 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7795275590551181,"balanced_accuracy":0.5512265512265512,"precision":0.5,"recall":0.14285714285714285,"f1":0.2222222222222222,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"Claude Opus 4.8, high","accuracy":0.37168141592920356,"balanced_accuracy":0.33576480990274093,"precision":0.11864406779661017,"recall":0.2692307692307692,"f1":0.16470588235294117,"valid_fraction":0.889763779527559,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.5589,"accuracy":0.7937,"f1":0.2353,"precision":0.5714,"recall":0.1481,"n_samples":127,"tp":4,"tn":96,"fp":3,"fn":23,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","accuracy":0.8031496062992126,"balanced_accuracy":0.5535714285714286,"precision":1,"recall":0.10714285714285714,"f1":0.1935483870967742,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.43019480519480524,"accuracy":0.29133858267716534,"f1":0.296875,"precision":0.19,"recall":0.6785714285714286,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy":0.5631313131313131,"accuracy":0.7946428571428571,"f1":0.25806451612903225,"precision":0.4444444444444444,"recall":0.18181818181818182,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8818897637795275,"rank":1}],"name":"Phase 1 - 2025Q2","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-1-2025q3":{"meta":{"title":"ClinBench Quarterly: Phase 1 - 2025Q3","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q3. N=30 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.5666666666666667,"balanced_accuracy":0.562200956937799,"precision":0.42857142857142855,"recall":0.5454545454545454,"f1":0.48,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Claude Opus 4.8, high","accuracy":0.6363636363636364,"balanced_accuracy":0.6363636363636364,"precision":0.6363636363636364,"recall":0.6363636363636364,"f1":0.6363636363636364,"valid_fraction":0.7333333333333333,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.5885,"accuracy":0.6,"f1":0.5,"precision":0.4615,"recall":0.5455,"n_samples":30,"tp":6,"tn":12,"fp":7,"fn":5,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy":0.6,"balanced_accuracy":0.569377990430622,"precision":0.45454545454545453,"recall":0.45454545454545453,"f1":0.45454545454545453,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.5550239234449761,"accuracy":0.5333333333333333,"f1":0.5,"precision":0.4117647058823529,"recall":0.6363636363636364,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","balanced_accuracy":0.6148325358851674,"accuracy":0.6333333333333333,"f1":0.5217391304347826,"precision":0.5,"recall":0.5454545454545454,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":2}],"name":"Phase 1 - 2025Q3","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-1-2025q4":{"meta":{"title":"ClinBench Quarterly: Phase 1 - 2025Q4","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2025Q4. N=23 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.5652173913043478,"balanced_accuracy":0.5436507936507937,"precision":0.4444444444444444,"recall":0.4444444444444444,"f1":0.4444444444444444,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Claude Opus 4.8, high","accuracy":0.5652173913043478,"balanced_accuracy":0.5634920634920635,"precision":0.45454545454545453,"recall":0.5555555555555556,"f1":0.5,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.4524,"accuracy":0.4783,"f1":0.3333,"precision":0.3333,"recall":0.3333,"n_samples":23,"tp":3,"tn":8,"fp":6,"fn":6,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.5652173913043478,"balanced_accuracy":0.5436507936507937,"precision":0.4444444444444444,"recall":0.4444444444444444,"f1":0.4444444444444444,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.4722222222222222,"accuracy":0.4782608695652174,"f1":0.4,"precision":0.36363636363636365,"recall":0.4444444444444444,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy":0.7817460317460317,"accuracy":0.782608695652174,"f1":0.7368421052631579,"precision":0.7,"recall":0.7777777777777778,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1}],"name":"Phase 1 - 2025Q4","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-1-2026q1":{"meta":{"title":"ClinBench Quarterly: Phase 1 - 2026Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 1) for trials with results submitted in 2026Q1. N=3 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":1,"balanced_accuracy":1,"precision":0,"recall":0,"f1":0,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Claude Opus 4.8, high","accuracy":1,"balanced_accuracy":1,"precision":0,"recall":0,"f1":0,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":1,"accuracy":1,"f1":0,"precision":0,"recall":0,"n_samples":3,"tp":0,"tn":3,"fp":0,"fn":0,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","accuracy":1,"balanced_accuracy":1,"precision":0,"recall":0,"f1":0,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"GPT 5.6 Sol, high","balanced_accuracy":1,"accuracy":1,"f1":0,"precision":0,"recall":0,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy":1,"accuracy":1,"f1":0,"precision":0,"recall":0,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1}],"name":"Phase 1 - 2026Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-2-2025q1":{"meta":{"title":"ClinBench Quarterly: Phase 2 - 2025Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q1. N=146 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.821917808219178,"balanced_accuracy":0.8198924731182795,"precision":0.78125,"recall":0.8064516129032258,"f1":0.7936507936507936,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8110236220472441,"balanced_accuracy":0.812280701754386,"precision":0.7704918032786885,"recall":0.8245614035087719,"f1":0.7966101694915254,"valid_fraction":0.8698630136986302,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7897,"accuracy":0.7945,"f1":0.7581,"precision":0.7581,"recall":0.7581,"n_samples":146,"tp":47,"tn":69,"fp":15,"fn":15,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.7945205479452054,"balanced_accuracy":0.7918586789554531,"precision":0.75,"recall":0.7741935483870968,"f1":0.7619047619047619,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8045314900153611,"accuracy":0.7945205479452054,"f1":0.782608695652174,"precision":0.7105263157894737,"recall":0.8709677419354839,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.8233260912502469,"accuracy":0.8263888888888888,"f1":0.7967479674796748,"precision":0.7903225806451613,"recall":0.8032786885245902,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.9863013698630136,"rank":1}],"name":"Phase 2 - 2025Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-2-2025q2":{"meta":{"title":"ClinBench Quarterly: Phase 2 - 2025Q2","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q2. N=163 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.852760736196319,"balanced_accuracy":0.8537290715372907,"precision":0.8837209302325582,"recall":0.8444444444444444,"f1":0.8636363636363636,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"Claude Opus 4.8, high","accuracy":0.7971014492753623,"balanced_accuracy":0.7943368107302533,"precision":0.8181818181818182,"recall":0.8181818181818182,"f1":0.8181818181818182,"valid_fraction":0.8466257668711656,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8674,"accuracy":0.865,"f1":0.8736,"precision":0.9048,"recall":0.8444,"n_samples":163,"tp":76,"tn":65,"fp":8,"fn":14,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","accuracy":0.8159509202453987,"balanced_accuracy":0.8242770167427702,"precision":0.9054054054054054,"recall":0.7444444444444445,"f1":0.8170731707317073,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8729832572298326,"accuracy":0.8711656441717791,"f1":0.88,"precision":0.9058823529411765,"recall":0.8555555555555555,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy":0.8576103500761034,"accuracy":0.852760736196319,"f1":0.8588235294117647,"precision":0.9125,"recall":0.8111111111111111,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":3}],"name":"Phase 2 - 2025Q2","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-2-2025q3":{"meta":{"title":"ClinBench Quarterly: Phase 2 - 2025Q3","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q3. N=145 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8896551724137931,"balanced_accuracy":0.8179535467671061,"precision":0.7037037037037037,"recall":0.7037037037037037,"f1":0.7037037037037037,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.864,"balanced_accuracy":0.8148795648795648,"precision":0.6551724137931034,"recall":0.7307692307692307,"f1":0.6909090909090909,"valid_fraction":0.8620689655172413,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7354,"accuracy":0.8483,"f1":0.5769,"precision":0.6,"recall":0.5556,"n_samples":145,"tp":15,"tn":108,"fp":10,"fn":12,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.8827586206896552,"balanced_accuracy":0.7565913370998116,"precision":0.75,"recall":0.5555555555555556,"f1":0.6382978723404256,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.7766792215944758,"accuracy":0.8689655172413793,"f1":0.6415094339622641,"precision":0.6538461538461539,"recall":0.6296296296296297,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.8630636604774535,"accuracy":0.9225352112676056,"f1":0.7843137254901961,"precision":0.8,"recall":0.7692307692307693,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.9793103448275862,"rank":1}],"name":"Phase 2 - 2025Q3","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-2-2025q4":{"meta":{"title":"ClinBench Quarterly: Phase 2 - 2025Q4","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2025Q4. N=114 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7368421052631579,"balanced_accuracy":0.7266624040920716,"precision":0.6739130434782609,"recall":0.6739130434782609,"f1":0.6739130434782609,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"Claude Opus 4.8, high","accuracy":0.7623762376237624,"balanced_accuracy":0.7419807538091419,"precision":0.7878787878787878,"recall":0.6046511627906976,"f1":0.6842105263157895,"valid_fraction":0.8859649122807017,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7634,"accuracy":0.7807,"f1":0.7126,"precision":0.7561,"recall":0.6739,"n_samples":114,"tp":31,"tn":58,"fp":10,"fn":15,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","accuracy":0.7982456140350878,"balanced_accuracy":0.7640664961636829,"precision":0.8709677419354839,"recall":0.5869565217391305,"f1":0.7012987012987013,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.7746163682864451,"accuracy":0.7982456140350878,"f1":0.7228915662650602,"precision":0.8108108108108109,"recall":0.6521739130434783,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","balanced_accuracy":0.7825174825174825,"accuracy":0.7978723404255319,"f1":0.7397260273972602,"precision":0.7941176470588235,"recall":0.6923076923076923,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8245614035087719,"rank":1}],"name":"Phase 2 - 2025Q4","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-2-2026q1":{"meta":{"title":"ClinBench Quarterly: Phase 2 - 2026Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 2) for trials with results submitted in 2026Q1. N=40 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.825,"balanced_accuracy":0.825,"precision":0.9333333333333333,"recall":0.7,"f1":0.8,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Claude Opus 4.8, high","accuracy":0.8125,"balanced_accuracy":0.807843137254902,"precision":0.8461538461538461,"recall":0.7333333333333333,"f1":0.7857142857142857,"valid_fraction":0.8,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.75,"accuracy":0.75,"f1":0.7222,"precision":0.8125,"recall":0.65,"n_samples":40,"tp":13,"tn":17,"fp":3,"fn":7,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy":0.775,"balanced_accuracy":0.7749999999999999,"precision":0.9230769230769231,"recall":0.6,"f1":0.7272727272727273,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.75,"accuracy":0.75,"f1":0.7368421052631579,"precision":0.7777777777777778,"recall":0.7,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","balanced_accuracy":0.825,"accuracy":0.825,"f1":0.8108108108108109,"precision":0.8823529411764706,"recall":0.75,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1}],"name":"Phase 2 - 2026Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-3-2025q1":{"meta":{"title":"ClinBench Quarterly: Phase 3 - 2025Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q1. N=119 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.907563025210084,"balanced_accuracy":0.9123376623376623,"precision":0.9583333333333334,"recall":0.8961038961038961,"f1":0.9261744966442953,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8532110091743119,"balanced_accuracy":0.8472826086956522,"precision":0.8955223880597015,"recall":0.8695652173913043,"f1":0.8823529411764706,"valid_fraction":0.9159663865546218,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8929,"accuracy":0.8824,"f1":0.9041,"precision":0.9565,"recall":0.8571,"n_samples":119,"tp":66,"tn":39,"fp":3,"fn":11,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy":0.8487394957983193,"balanced_accuracy":0.8398268398268398,"precision":0.8933333333333333,"recall":0.8701298701298701,"f1":0.881578947368421,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8896103896103895,"accuracy":0.8991596638655462,"f1":0.922077922077922,"precision":0.922077922077922,"recall":0.922077922077922,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.9147619047619048,"accuracy":0.8974358974358975,"f1":0.9142857142857143,"precision":0.9846153846153847,"recall":0.8533333333333334,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.9831932773109243,"rank":1}],"name":"Phase 3 - 2025Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-3-2025q2":{"meta":{"title":"ClinBench Quarterly: Phase 3 - 2025Q2","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q2. N=133 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8947368421052632,"balanced_accuracy":0.8918072289156627,"precision":0.9259259259259259,"recall":0.9036144578313253,"f1":0.9146341463414634,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Claude Opus 4.8, high","accuracy":0.875,"balanced_accuracy":0.8717909996979765,"precision":0.918918918918919,"recall":0.8831168831168831,"f1":0.9006622516556292,"valid_fraction":0.9022556390977443,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8777,"accuracy":0.8722,"f1":0.8931,"precision":0.9342,"recall":0.8554,"n_samples":133,"tp":71,"tn":45,"fp":5,"fn":12,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","accuracy":0.8721804511278195,"balanced_accuracy":0.8657831325301204,"precision":0.9024390243902439,"recall":0.891566265060241,"f1":0.896969696969697,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.9098795180722892,"accuracy":0.9172932330827067,"f1":0.9341317365269461,"precision":0.9285714285714286,"recall":0.9397590361445783,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy":0.909767718880286,"accuracy":0.8991596638655462,"f1":0.9130434782608695,"precision":0.9692307692307692,"recall":0.863013698630137,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8947368421052632,"rank":2}],"name":"Phase 3 - 2025Q2","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-3-2025q3":{"meta":{"title":"ClinBench Quarterly: Phase 3 - 2025Q3","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q3. N=100 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.84,"balanced_accuracy":0.8327205882352942,"precision":0.90625,"recall":0.8529411764705882,"f1":0.8787878787878788,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"Claude Opus 4.8, high","accuracy":0.8295454545454546,"balanced_accuracy":0.8224137931034483,"precision":0.8909090909090909,"recall":0.8448275862068966,"f1":0.8672566371681416,"valid_fraction":0.88,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8493,"accuracy":0.84,"f1":0.875,"precision":0.9333,"recall":0.8235,"n_samples":100,"tp":56,"tn":28,"fp":4,"fn":12,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","accuracy":0.89,"balanced_accuracy":0.9108455882352942,"precision":0.9830508474576272,"recall":0.8529411764705882,"f1":0.9133858267716536,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.8630514705882353,"accuracy":0.87,"f1":0.9022556390977443,"precision":0.9230769230769231,"recall":0.8823529411764706,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","balanced_accuracy":0.9208333333333334,"accuracy":0.9069767441860465,"f1":0.9245283018867925,"precision":0.98,"recall":0.875,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.86,"rank":1}],"name":"Phase 3 - 2025Q3","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-3-2025q4":{"meta":{"title":"ClinBench Quarterly: Phase 3 - 2025Q4","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2025Q4. N=113 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8938053097345132,"balanced_accuracy":0.8956639566395663,"precision":0.9411764705882353,"recall":0.8888888888888888,"f1":0.9142857142857143,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":4},{"model":"Claude Opus 4.8, high","accuracy":0.8979591836734694,"balanced_accuracy":0.9081585081585082,"precision":0.9661016949152542,"recall":0.8769230769230769,"f1":0.9193548387096774,"valid_fraction":0.8672566371681416,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8748,"accuracy":0.8673,"f1":0.8905,"precision":0.9385,"recall":0.8472,"n_samples":113,"tp":61,"tn":37,"fp":4,"fn":11,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.8849557522123894,"balanced_accuracy":0.893970189701897,"precision":0.9538461538461539,"recall":0.8611111111111112,"f1":0.9051094890510949,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.9286924119241193,"accuracy":0.9292035398230089,"f1":0.9436619718309859,"precision":0.9571428571428572,"recall":0.9305555555555556,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy":0.9034240561896401,"accuracy":0.8910891089108911,"f1":0.9133858267716536,"precision":0.9666666666666667,"recall":0.8656716417910447,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.8938053097345132,"rank":3}],"name":"Phase 3 - 2025Q4","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-3-2026q1":{"meta":{"title":"ClinBench Quarterly: Phase 3 - 2026Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 3) for trials with results submitted in 2026Q1. N=38 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.868421052631579,"balanced_accuracy":0.9264705882352942,"precision":1,"recall":0.8529411764705882,"f1":0.9206349206349206,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8787878787878788,"balanced_accuracy":0.8232758620689655,"precision":0.9629629629629629,"recall":0.896551724137931,"f1":0.9285714285714286,"valid_fraction":0.868421052631579,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7574,"accuracy":0.7632,"f1":0.8525,"precision":0.963,"recall":0.7647,"n_samples":38,"tp":26,"tn":3,"fp":1,"fn":8,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy":0.7894736842105263,"balanced_accuracy":0.7720588235294117,"precision":0.9642857142857143,"recall":0.7941176470588235,"f1":0.8709677419354839,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.7867647058823529,"accuracy":0.8157894736842105,"f1":0.8888888888888888,"precision":0.9655172413793104,"recall":0.8235294117647058,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.9411764705882353,"accuracy":0.8947368421052632,"f1":0.9375,"precision":1,"recall":0.8823529411764706,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1}],"name":"Phase 3 - 2026Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-4-2025q1":{"meta":{"title":"ClinBench Quarterly: Phase 4 - 2025Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q1. N=35 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7714285714285715,"balanced_accuracy":0.7857142857142857,"precision":0.6666666666666666,"recall":0.8571428571428571,"f1":0.75,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Claude Opus 4.8, high","accuracy":0.7714285714285715,"balanced_accuracy":0.7857142857142857,"precision":0.6666666666666666,"recall":0.8571428571428571,"f1":0.75,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.9048,"accuracy":0.9143,"f1":0.8889,"precision":0.9231,"recall":0.8571,"n_samples":35,"tp":12,"tn":20,"fp":1,"fn":2,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","accuracy":0.8,"balanced_accuracy":0.7619047619047619,"precision":0.8888888888888888,"recall":0.5714285714285714,"f1":0.6956521739130435,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.75,"accuracy":0.7428571428571429,"f1":0.7096774193548387,"precision":0.6470588235294118,"recall":0.7857142857142857,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","balanced_accuracy":0.8333333333333333,"accuracy":0.8285714285714286,"f1":0.8,"precision":0.75,"recall":0.8571428571428571,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":2}],"name":"Phase 4 - 2025Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-4-2025q2":{"meta":{"title":"ClinBench Quarterly: Phase 4 - 2025Q2","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q2. N=31 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8064516129032258,"balanced_accuracy":0.8020833333333333,"precision":0.75,"recall":0.9375,"f1":0.8333333333333334,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8709677419354839,"balanced_accuracy":0.86875,"precision":0.8333333333333334,"recall":0.9375,"f1":0.8823529411764706,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7417,"accuracy":0.7419,"f1":0.75,"precision":0.75,"recall":0.75,"n_samples":31,"tp":12,"tn":11,"fp":4,"fn":4,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy":0.7096774193548387,"balanced_accuracy":0.7145833333333333,"precision":0.8181818181818182,"recall":0.5625,"f1":0.6666666666666666,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.76875,"accuracy":0.7741935483870968,"f1":0.8108108108108109,"precision":0.7142857142857143,"recall":0.9375,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.7946428571428572,"accuracy":0.8,"f1":0.8235294117647058,"precision":0.7777777777777778,"recall":0.875,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0.4838709677419355,"rank":3}],"name":"Phase 4 - 2025Q2","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-4-2025q3":{"meta":{"title":"ClinBench Quarterly: Phase 4 - 2025Q3","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q3. N=22 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7272727272727273,"balanced_accuracy":0.6857142857142857,"precision":0.5714285714285714,"recall":0.5714285714285714,"f1":0.5714285714285714,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Claude Opus 4.8, high","accuracy":0.7272727272727273,"balanced_accuracy":0.6857142857142857,"precision":0.5714285714285714,"recall":0.5714285714285714,"f1":0.5714285714285714,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.719,"accuracy":0.7727,"f1":0.6154,"precision":0.6667,"recall":0.5714,"n_samples":22,"tp":4,"tn":13,"fp":2,"fn":3,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","accuracy":0.7272727272727273,"balanced_accuracy":0.6857142857142857,"precision":0.5714285714285714,"recall":0.5714285714285714,"f1":0.5714285714285714,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":3},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.6523809523809523,"accuracy":0.6818181818181818,"f1":0.5333333333333333,"precision":0.5,"recall":0.5714285714285714,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","balanced_accuracy":0.719047619047619,"accuracy":0.7727272727272727,"f1":0.6153846153846154,"precision":0.6666666666666666,"recall":0.5714285714285714,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1}],"name":"Phase 4 - 2025Q3","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-4-2025q4":{"meta":{"title":"ClinBench Quarterly: Phase 4 - 2025Q4","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2025Q4. N=43 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.6744186046511628,"balanced_accuracy":0.666304347826087,"precision":0.6666666666666666,"recall":0.782608695652174,"f1":0.72,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":6},{"model":"Claude Opus 4.8, high","accuracy":0.8108108108108109,"balanced_accuracy":0.8073529411764706,"precision":0.8095238095238095,"recall":0.85,"f1":0.8292682926829268,"valid_fraction":0.8604651162790697,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":2},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.7196,"accuracy":0.7209,"f1":0.7391,"precision":0.7391,"recall":0.7391,"n_samples":43,"tp":17,"tn":14,"fp":6,"fn":6,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy":0.6976744186046512,"balanced_accuracy":0.7076086956521739,"precision":0.8125,"recall":0.5652173913043478,"f1":0.6666666666666666,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":5},{"model":"GPT 5.6 Sol, high","balanced_accuracy":0.716304347826087,"accuracy":0.7209302325581395,"f1":0.75,"precision":0.72,"recall":0.782608695652174,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","balanced_accuracy":0.8565217391304347,"accuracy":0.8604651162790697,"f1":0.875,"precision":0.84,"recall":0.9130434782608695,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1}],"name":"Phase 4 - 2025Q4","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-4-2026q1":{"meta":{"title":"ClinBench Quarterly: Phase 4 - 2026Q1","category":"Clinical Trials > ClinBench Quarterly","description":"Clinical trial outcome prediction (Phase 4) for trials with results submitted in 2026Q1. N=6 samples.","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["balanced_accuracy","accuracy","f1","precision","recall","n_samples","tp","tn","fp","fn","valid_fraction"],"leaderboard_metric":"balanced_accuracy","results":[{"model":"GPT 5.5, medium","accuracy":1,"balanced_accuracy":1,"precision":0,"recall":0,"f1":0,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Claude Opus 4.8, high","accuracy":1,"balanced_accuracy":1,"precision":0,"recall":0,"f1":0,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"Gemini 3.1 Pro, high","balanced_accuracy":0.8333,"accuracy":0.8333,"f1":0,"precision":0,"recall":0,"n_samples":6,"tp":0,"tn":5,"fp":1,"fn":0,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy":1,"balanced_accuracy":1,"precision":0,"recall":0,"f1":0,"valid_fraction":1,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"rank":1},{"model":"GPT 5.6 Sol, high","balanced_accuracy":1,"accuracy":1,"f1":0,"precision":0,"recall":0,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","balanced_accuracy":null,"accuracy":null,"f1":null,"precision":null,"recall":null,"n_samples":null,"tp":null,"tn":null,"fp":null,"fn":null,"valid_fraction":0,"rank":6}],"name":"Phase 4 - 2026Q1","metric_info":{"balanced_accuracy":{"label":"Balanced Accuracy","shortLabel":"BAcc","direction":"higher_better","description":"Average of sensitivity and specificity, robust to class imbalance","units":"ratio"},"accuracy":{"label":"Accuracy","shortLabel":"Acc","direction":"higher_better","description":"Proportion of correct predictions","units":"ratio"},"precision":{"label":"Precision","shortLabel":"Prec","direction":"higher_better","description":"Positive predictive value","units":"ratio"},"recall":{"label":"Recall","shortLabel":"Rec","direction":"higher_better","description":"Sensitivity - proportion of actual successes correctly predicted","units":"ratio"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"},"n_samples":{"label":"Sample Count","shortLabel":"N","direction":"higher_better","description":"Number of test samples in this quarter"},"tp":{"label":"True Positives","shortLabel":"TP","direction":"higher_better","description":"Correctly predicted successes"},"tn":{"label":"True Negatives","shortLabel":"TN","direction":"higher_better","description":"Correctly predicted failures"},"fp":{"label":"False Positives","shortLabel":"FP","direction":"lower_better","description":"Incorrectly predicted as success"},"fn":{"label":"False Negatives","shortLabel":"FN","direction":"lower_better","description":"Incorrectly predicted as failure"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"cli-clinbench-quarterly-phase-1":{"meta":{"title":"ClinBench Quarterly: Phase 1","category":"Clinical Trials > ClinBench Quarterly","description":"outcome_phase1_xl","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["accuracy","balanced_accuracy","precision","recall","f1","valid_fraction"],"leaderboard_metric":"accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7531380753138075,"balanced_accuracy":0.6591847265221877,"precision":0.5882352941176471,"recall":0.4411764705882353,"f1":0.5042016806722689,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.47368421052631576,"balanced_accuracy":0.4645547186530793,"precision":0.2621359223300971,"recall":0.4426229508196721,"f1":0.32926829268292684,"valid_fraction":0.8744769874476988,"rank":6},{"model":"Gemini 3.1 Pro, high","accuracy":0.7531380753138075,"balanced_accuracy":0.6370399036807706,"precision":0.6097560975609756,"recall":0.36764705882352944,"f1":0.45871559633027525,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","accuracy":0.7322175732217573,"balanced_accuracy":0.6312779497764017,"precision":0.54,"recall":0.39705882352941174,"f1":0.4576271186440678,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","accuracy":0.4895397489539749,"balanced_accuracy":0.5458376332989336,"precision":0.3150684931506849,"recall":0.6764705882352942,"f1":0.42990654205607476,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","accuracy":0.7639484978540773,"balanced_accuracy":0.6830156446682252,"precision":0.6111111111111112,"recall":0.4925373134328358,"f1":0.5454545454545454,"valid_fraction":0.9748953974895398,"rank":1}],"name":"Phase 1","metric_info":{"accuracy":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"precision":{"label":"precision_score","shortLabel":"PRECI","direction":"higher_better"},"recall":{"label":"recall_score","shortLabel":"RECAL","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"}}},"cli-clinbench-quarterly-phase-2":{"meta":{"title":"ClinBench Quarterly: Phase 2","category":"Clinical Trials > ClinBench Quarterly","description":"outcome_phase2_xl","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["accuracy","balanced_accuracy","precision","recall","f1","valid_fraction"],"leaderboard_metric":"accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8240131578947368,"balanced_accuracy":0.8167931635464103,"precision":0.7827868852459017,"recall":0.7795918367346939,"f1":0.7811860940695297,"valid_fraction":1,"rank":3},{"model":"Claude Opus 4.8, high","accuracy":0.8090737240075614,"balanced_accuracy":0.7977608598298254,"precision":0.7684729064039408,"recall":0.7428571428571429,"f1":0.7554479418886199,"valid_fraction":0.8700657894736842,"rank":6},{"model":"Gemini 3.1 Pro, high","accuracy":0.8322368421052632,"balanced_accuracy":0.8143925338730533,"precision":0.8388625592417062,"recall":0.7224489795918367,"f1":0.7763157894736842,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","accuracy":0.8141447368421053,"balanced_accuracy":0.8018946421543824,"precision":0.7869565217391304,"recall":0.7387755102040816,"f1":0.7621052631578947,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","accuracy":0.8223684210526315,"balanced_accuracy":0.8160791589363018,"precision":0.7773279352226721,"recall":0.7836734693877551,"f1":0.7804878048780488,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy":0.8520408163265306,"balanced_accuracy":0.8370117927079952,"precision":0.8571428571428571,"recall":0.759493670886076,"f1":0.8053691275167785,"valid_fraction":0.9671052631578947,"rank":1}],"name":"Phase 2","metric_info":{"accuracy":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"precision":{"label":"precision_score","shortLabel":"PRECI","direction":"higher_better"},"recall":{"label":"recall_score","shortLabel":"RECAL","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"}}},"cli-clinbench-quarterly-phase-3":{"meta":{"title":"ClinBench Quarterly: Phase 3","category":"Clinical Trials > ClinBench Quarterly","description":"outcome_phase3_xl","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["accuracy","balanced_accuracy","precision","recall","f1","valid_fraction"],"leaderboard_metric":"accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.8986083499005965,"balanced_accuracy":0.9002675123126528,"precision":0.9492063492063492,"recall":0.8952095808383234,"f1":0.9214175654853621,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.8674157303370786,"balanced_accuracy":0.8589080459770115,"precision":0.916955017301038,"recall":0.8833333333333333,"f1":0.8998302207130731,"valid_fraction":0.8846918489065606,"rank":4},{"model":"Gemini 3.1 Pro, high","accuracy":0.8667992047713717,"balanced_accuracy":0.8865464337596995,"precision":0.968421052631579,"recall":0.8263473053892215,"f1":0.8917609046849758,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy":0.8628230616302187,"balanced_accuracy":0.8630903872727917,"precision":0.9260450160771704,"recall":0.8622754491017964,"f1":0.8930232558139535,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","accuracy":0.9005964214711729,"balanced_accuracy":0.9017645183006767,"precision":0.9493670886075949,"recall":0.8982035928143712,"f1":0.9230769230769231,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","accuracy":0.891170431211499,"balanced_accuracy":0.9089518991165144,"precision":0.9787234042553191,"recall":0.8544891640866873,"f1":0.912396694214876,"valid_fraction":0.9681908548707754,"rank":3}],"name":"Phase 3","metric_info":{"accuracy":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"precision":{"label":"precision_score","shortLabel":"PRECI","direction":"higher_better"},"recall":{"label":"recall_score","shortLabel":"RECAL","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"}}},"cli-clinbench-quarterly-phase-4":{"meta":{"title":"ClinBench Quarterly: Phase 4","category":"Clinical Trials > ClinBench Quarterly","description":"outcome_phase4_xl","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["accuracy","balanced_accuracy","precision","recall","f1","valid_fraction"],"leaderboard_metric":"accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.7664233576642335,"balanced_accuracy":0.771969696969697,"precision":0.7,"recall":0.8166666666666667,"f1":0.7538461538461538,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy":0.7903225806451613,"balanced_accuracy":0.7905926122774383,"precision":0.7368421052631579,"recall":0.7924528301886793,"f1":0.7636363636363637,"valid_fraction":0.9051094890510949,"rank":2},{"model":"Gemini 3.1 Pro, high","accuracy":0.7883211678832117,"balanced_accuracy":0.7822510822510822,"precision":0.7719298245614035,"recall":0.7333333333333333,"f1":0.7521367521367521,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy":0.7664233576642335,"balanced_accuracy":0.7498917748917748,"precision":0.8043478260869565,"recall":0.6166666666666667,"f1":0.6981132075471698,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","accuracy":0.7591240875912408,"balanced_accuracy":0.7673160173160174,"precision":0.684931506849315,"recall":0.8333333333333334,"f1":0.7518796992481203,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","accuracy":0.8454545454545455,"balanced_accuracy":0.8486115757778521,"precision":0.7962962962962963,"recall":0.8775510204081632,"f1":0.8349514563106796,"valid_fraction":0.8029197080291971,"rank":1}],"name":"Phase 4","metric_info":{"accuracy":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"precision":{"label":"precision_score","shortLabel":"PRECI","direction":"higher_better"},"recall":{"label":"recall_score","shortLabel":"RECAL","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"}}},"cli-clinbench-quarterly-all-phases":{"meta":{"title":"ClinBench Quarterly: All Phases","category":"Clinical Trials > ClinBench Quarterly","description":"outcome_all_xl","metric_types":[],"paper_link":"#","source_file":"clinbench_arena_quarterly_ci.json","original_category":"Clinical Trials"},"radar_keys":["accuracy","balanced_accuracy","precision","recall","f1","valid_fraction"],"leaderboard_metric":"accuracy","results":[{"model":"GPT 5.5, medium","accuracy":0.839946200403497,"balanced_accuracy":0.8383019620643384,"precision":0.8505231689088192,"recall":0.8048090523338048,"f1":0.8270348837209303,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy":0.7913779830638953,"balanced_accuracy":0.7912678062678062,"precision":0.7797147385103012,"recall":0.7884615384615384,"f1":0.7840637450199203,"valid_fraction":0.8735709482178884,"rank":6},{"model":"Gemini 3.1 Pro, high","accuracy":0.8285137861466039,"balanced_accuracy":0.8244260689805244,"precision":0.8791946308724832,"recall":0.7411598302687411,"f1":0.8042977743668457,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy":0.816408876933423,"balanced_accuracy":0.8131523591919632,"precision":0.8488745980707395,"recall":0.7468175388967468,"f1":0.7945823927765236,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","accuracy":0.7955615332885003,"balanced_accuracy":0.7974504043810975,"precision":0.7586649550706034,"recall":0.8359264497878359,"f1":0.7954239569313594,"valid_fraction":1,"rank":5},{"model":"Claude Opus 5, high","accuracy":0.8596363636363636,"balanced_accuracy":0.8573712922067462,"precision":0.893581081081081,"recall":0.802731411229135,"f1":0.8457234212629896,"valid_fraction":0.9246805648957633,"rank":1}],"name":"All Phases","metric_info":{"accuracy":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"precision":{"label":"precision_score","shortLabel":"PRECI","direction":"higher_better"},"recall":{"label":"recall_score","shortLabel":"RECAL","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"},"f1":{"label":"F1 Score","shortLabel":"F1","direction":"higher_better","description":"Harmonic mean of precision and recall","units":"ratio"}}},"lon-longevitybench-peachy-geo-dnam-age-pairwise-lb-0006":{"meta":{"title":"LongevityBench (Peachy): GEO DNAm Age / Pairwise (LB-0006)","category":"Longevity > LongevityBench (Peachy)","description":"Built from public GEO DNA-methylation datasets (Illumina arrays). Two blood donors are each described by the methylation levels at their most age-informative CpG sites (hundreds of sites, with gene annotations); the model decides which donor is older.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7778,"balanced_accuracy_score":0.7787,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy_score":0.6667,"balanced_accuracy_score":0.665,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7111,"balanced_accuracy_score":0.7144,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy_score":0.7555555555555555,"balanced_accuracy_score":0.7549407114624507,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7555555555555555,"balanced_accuracy_score":0.7539525691699605,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","accuracy_score":0.8108108108108109,"balanced_accuracy_score":0.8029411764705883,"valid_fraction":0.8222222222222222,"rank":1}],"name":"GEO DNAm Age / Pairwise (LB-0006)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-gtex-transcriptomic-age-pairwise-lb-0018":{"meta":{"title":"LongevityBench (Peachy): GTEx Transcriptomic Age / Pairwise (LB-0018)","category":"Longevity > LongevityBench (Peachy)","description":"Built from the GTEx project (post-mortem tissue RNA-seq). The two samples are compared through the ~100 genes whose expression differs most between them, and the model predicts which donor was older.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.4889,"balanced_accuracy_score":0.4901,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy_score":0.5111,"balanced_accuracy_score":0.5138,"valid_fraction":1,"rank":3},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.4667,"balanced_accuracy_score":0.4625,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy_score":0.5111111111111111,"balanced_accuracy_score":0.5108695652173914,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.4666666666666667,"balanced_accuracy_score":0.46739130434782605,"valid_fraction":1,"rank":6},{"model":"Claude Opus 5, high","accuracy_score":0.5555555555555556,"balanced_accuracy_score":0.5622529644268774,"valid_fraction":1,"rank":1}],"name":"GTEx Transcriptomic Age / Pairwise (LB-0018)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-nhanes-biomarkers-age-pairwise-lb-0034":{"meta":{"title":"LongevityBench (Peachy): NHANES Biomarkers Age / Pairwise (LB-0034)","category":"Longevity > LongevityBench (Peachy)","description":"Built from the U.S. NHANES health survey. Two people are described by routine clinical data, such as blood-panel biomarkers, body measurements and basic demographics. The model decides which is older.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7111,"balanced_accuracy_score":0.7105,"valid_fraction":1,"rank":5},{"model":"Claude Opus 4.8, high","accuracy_score":0.7111,"balanced_accuracy_score":0.7125,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7333,"balanced_accuracy_score":0.7312,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","accuracy_score":0.7111111111111111,"balanced_accuracy_score":0.7104743083003953,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7111111111111111,"balanced_accuracy_score":0.7094861660079052,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","accuracy_score":0.7333333333333333,"balanced_accuracy_score":0.7312252964426877,"valid_fraction":1,"rank":1}],"name":"NHANES Biomarkers Age / Pairwise (LB-0034)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-nhanes-biomarkers-mortality-binary-lb-0042":{"meta":{"title":"LongevityBench (Peachy): NHANES Biomarkers Mortality / Binary (LB-0042)","category":"Longevity > LongevityBench (Peachy)","description":"From one person's NHANES clinical and blood-biomarker profile, the model predicts whether they were still alive 10 years post-examination.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7333,"balanced_accuracy_score":0.7322,"valid_fraction":1,"rank":6},{"model":"Claude Opus 4.8, high","accuracy_score":0.7556,"balanced_accuracy_score":0.754,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7778,"balanced_accuracy_score":0.7777,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","accuracy_score":0.7777777777777778,"balanced_accuracy_score":0.7786561264822134,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7777777777777778,"balanced_accuracy_score":0.7776679841897234,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","accuracy_score":0.7777777777777778,"balanced_accuracy_score":0.7786561264822134,"valid_fraction":1,"rank":2}],"name":"NHANES Biomarkers Mortality / Binary (LB-0042)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-nhanes-biomarkers-mortality-pairwise-lb-0050":{"meta":{"title":"LongevityBench (Peachy): NHANES Biomarkers Mortality / Pairwise (LB-0050)","category":"Longevity > LongevityBench (Peachy)","description":"Given two people's NHANES clinical profiles, the model predicts which of them survived longer after the survey.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.6222,"balanced_accuracy_score":0.6215,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy_score":0.6222,"balanced_accuracy_score":0.6235,"valid_fraction":1,"rank":4},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.6,"balanced_accuracy_score":0.5998,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy_score":0.6444444444444445,"balanced_accuracy_score":0.6432806324110671,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.6222222222222222,"balanced_accuracy_score":0.6205533596837944,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","accuracy_score":0.65625,"balanced_accuracy_score":0.65625,"valid_fraction":0.7111111111111111,"rank":1}],"name":"NHANES Biomarkers Mortality / Pairwise (LB-0050)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-olink-proteomic-age-pairwise-lb-0066":{"meta":{"title":"LongevityBench (Peachy): Olink Proteomic Age / Pairwise (LB-0066)","category":"Longevity > LongevityBench (Peachy)","description":"Built from the Olink Explore 3072 plasma-proteomics panel (Kirsher et al., 2025). Two blood samples are compared through the proteins whose levels (NPX) differ most between them and the model decides which donor is older.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.9111,"balanced_accuracy_score":0.9101,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy_score":0.6667,"balanced_accuracy_score":0.6719,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","accuracy_score":1,"balanced_accuracy_score":1,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","accuracy_score":0.9555555555555556,"balanced_accuracy_score":0.9555335968379447,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.9111111111111111,"balanced_accuracy_score":0.9100790513833992,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","accuracy_score":0.8888888888888888,"balanced_accuracy_score":0.8893280632411067,"valid_fraction":1,"rank":5}],"name":"Olink Proteomic Age / Pairwise (LB-0066)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-opengenes-expression-binary_masked-lb-0086":{"meta":{"title":"LongevityBench (Peachy): OpenGenes Expression / binary_masked (LB-0086)","category":"Longevity > LongevityBench (Peachy)","description":"Built from the OpenGenes database. A gene's identity is masked and described purely by its properties: molecular functions, interaction partners and tissue expression; the model predicts whether its expression rises or falls with age. Hiding the name forces reasoning rather than recall.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7111,"balanced_accuracy_score":0.7065,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","accuracy_score":0.6889,"balanced_accuracy_score":0.6887,"valid_fraction":1,"rank":2},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.6,"balanced_accuracy_score":0.5958,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy_score":0.6,"balanced_accuracy_score":0.5958498023715415,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","accuracy_score":0.6222222222222222,"balanced_accuracy_score":0.6175889328063241,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy_score":0.6888888888888889,"balanced_accuracy_score":0.6877470355731226,"valid_fraction":1,"rank":3}],"name":"OpenGenes Expression / binary_masked (LB-0086)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-synergyage-lifespan-pairwise-ablated-lb-0122":{"meta":{"title":"LongevityBench (Peachy): SynergyAge Lifespan / Pairwise (ablated, LB-0122)","category":"Longevity > LongevityBench (Peachy)","description":"Built from the SynergyAge database of lifespan experiments (mostly the worm C. elegans). Given two multi-gene perturbations, the model predicts which one extends lifespan more. This is the 'minimal' variant, deliberately stripped of extra context.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.8846,"balanced_accuracy_score":0.8879,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy_score":0.7308,"balanced_accuracy_score":0.7545,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.9231,"balanced_accuracy_score":0.9212,"valid_fraction":1,"rank":1},{"model":"Grok 4.5, high","accuracy_score":0.7307692307692307,"balanced_accuracy_score":0.7424242424242424,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7692307692307693,"balanced_accuracy_score":0.7878787878787878,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy_score":0.8461538461538461,"balanced_accuracy_score":0.8545454545454545,"valid_fraction":1,"rank":3}],"name":"SynergyAge Lifespan / Pairwise (ablated, LB-0122)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-tcga-cancer-survival-pairwise-lb-0134":{"meta":{"title":"LongevityBench (Peachy): TCGA Cancer Survival / Pairwise (LB-0134)","category":"Longevity > LongevityBench (Peachy)","description":"Built from TCGA tumour transcriptomes. Two cancer patients are described by their tumour gene-activity and pathway profiles, and the model predicts which patient had the longer progression-free  survival.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","balanced_accuracy_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.7333,"balanced_accuracy_score":0.7362,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy_score":0.8444,"balanced_accuracy_score":0.8468,"valid_fraction":1,"rank":1},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.7333,"balanced_accuracy_score":0.7352,"valid_fraction":1,"rank":4},{"model":"Grok 4.5, high","accuracy_score":0.7777777777777778,"balanced_accuracy_score":0.7796442687747036,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.7555555555555555,"balanced_accuracy_score":0.7569169960474309,"valid_fraction":1,"rank":3}],"name":"TCGA Cancer Survival / Pairwise (LB-0134)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"balanced_accuracy_score":{"label":"Balanced accuracy","shortLabel":"BAcc","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-geo-dnam-age-multiclass-lb-0002":{"meta":{"title":"LongevityBench (Peachy): GEO DNAm Age / Multiclass (LB-0002)","category":"Longevity > LongevityBench (Peachy)","description":"From a GEO DNA-methylation profile, the model assigns the donor to one of 11 age decades","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","macro_f1_score","weighted_f1_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.3778,"macro_f1_score":0.3665,"weighted_f1_score":0.3637,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","accuracy_score":0.2889,"macro_f1_score":0.2596,"weighted_f1_score":0.261,"valid_fraction":1,"rank":3},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.2,"macro_f1_score":0.1892,"weighted_f1_score":0.1821,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","accuracy_score":0.37777777777777777,"macro_f1_score":0.3635773317591499,"weighted_f1_score":0.3743867243867243,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.28888888888888886,"macro_f1_score":0.24889100798191707,"weighted_f1_score":0.2564560131226798,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy_score":0.275,"macro_f1_score":0.25375180375180373,"weighted_f1_score":0.26434523809523813,"valid_fraction":0.8888888888888888,"rank":5}],"name":"GEO DNAm Age / Multiclass (LB-0002)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"macro_f1_score":{"label":"macro_f1_score","shortLabel":"MACRO","direction":"higher_better"},"weighted_f1_score":{"label":"weighted_f1_score","shortLabel":"WEIGH","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-gtex-transcriptomic-age-multiclass-lb-0014":{"meta":{"title":"LongevityBench (Peachy): GTEx Transcriptomic Age / Multiclass (LB-0014)","category":"Longevity > LongevityBench (Peachy)","description":"From a GTEx tissue expression profile, the model sorts the donor into one of six age groups.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","macro_f1_score","weighted_f1_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.3214,"macro_f1_score":0.2983,"weighted_f1_score":0.297,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","accuracy_score":0.2679,"macro_f1_score":0.2298,"weighted_f1_score":0.2325,"valid_fraction":1,"rank":3},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.2321,"macro_f1_score":0.2318,"weighted_f1_score":0.2337,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy_score":0.32142857142857145,"macro_f1_score":0.30963921556026813,"weighted_f1_score":0.30765096693198196,"valid_fraction":1,"rank":1},{"model":"GPT 5.6 Sol, high","accuracy_score":0.23214285714285715,"macro_f1_score":0.23816548948127894,"weighted_f1_score":0.2379310695570094,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy_score":0.22727272727272727,"macro_f1_score":0.2081807081807082,"weighted_f1_score":0.20174825174825176,"valid_fraction":0.7857142857142857,"rank":6}],"name":"GTEx Transcriptomic Age / Multiclass (LB-0014)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"macro_f1_score":{"label":"macro_f1_score","shortLabel":"MACRO","direction":"higher_better"},"weighted_f1_score":{"label":"weighted_f1_score","shortLabel":"WEIGH","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-nhanes-biomarkers-age-multiclass-lb-0030":{"meta":{"title":"LongevityBench (Peachy): NHANES Biomarkers Age / Multiclass (LB-0030)","category":"Longevity > LongevityBench (Peachy)","description":"From a person's NHANES clinical and blood measurements, the model places them into one of five age brackets.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","macro_f1_score","weighted_f1_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.4444,"macro_f1_score":0.4337,"weighted_f1_score":0.4337,"valid_fraction":1,"rank":3},{"model":"Claude Opus 4.8, high","accuracy_score":0.3778,"macro_f1_score":0.359,"weighted_f1_score":0.359,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.3778,"macro_f1_score":0.371,"weighted_f1_score":0.371,"valid_fraction":1,"rank":5},{"model":"Grok 4.5, high","accuracy_score":0.5111111111111111,"macro_f1_score":0.5030570409982175,"weighted_f1_score":0.5030570409982175,"valid_fraction":1,"rank":2},{"model":"GPT 5.6 Sol, high","accuracy_score":0.4222222222222222,"macro_f1_score":0.41259436469962785,"weighted_f1_score":0.4125943646996279,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","accuracy_score":0.53125,"macro_f1_score":0.5027272727272727,"weighted_f1_score":0.5072443181818181,"valid_fraction":0.7111111111111111,"rank":1}],"name":"NHANES Biomarkers Age / Multiclass (LB-0030)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"macro_f1_score":{"label":"macro_f1_score","shortLabel":"MACRO","direction":"higher_better"},"weighted_f1_score":{"label":"weighted_f1_score","shortLabel":"WEIGH","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-nhanes-biomarkers-mortality-multiclass-lb-0046":{"meta":{"title":"LongevityBench (Peachy): NHANES Biomarkers Mortality / Multiclass (LB-0046)","category":"Longevity > LongevityBench (Peachy)","description":"From a person's NHANES profile, the model predicts how long they survived after the survey, sorted into four time-to-event bands.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["accuracy_score","macro_f1_score","weighted_f1_score","valid_fraction"],"leaderboard_metric":"accuracy_score","results":[{"model":"GPT 5.5, medium","accuracy_score":0.4444,"macro_f1_score":0.441,"weighted_f1_score":0.4428,"valid_fraction":1,"rank":4},{"model":"Claude Opus 4.8, high","accuracy_score":0.4444,"macro_f1_score":0.4387,"weighted_f1_score":0.4409,"valid_fraction":1,"rank":4},{"model":"Gemini 3.1 Pro, high","accuracy_score":0.4667,"macro_f1_score":0.4627,"weighted_f1_score":0.4608,"valid_fraction":1,"rank":3},{"model":"Grok 4.5, high","accuracy_score":0.4222222222222222,"macro_f1_score":0.4196969696969697,"weighted_f1_score":0.41703703703703704,"valid_fraction":1,"rank":6},{"model":"GPT 5.6 Sol, high","accuracy_score":0.4888888888888889,"macro_f1_score":0.4867424242424242,"weighted_f1_score":0.488047138047138,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","accuracy_score":0.5111111111111111,"macro_f1_score":0.5076086956521739,"weighted_f1_score":0.5074396135265701,"valid_fraction":1,"rank":1}],"name":"NHANES Biomarkers Mortality / Multiclass (LB-0046)","metric_info":{"accuracy_score":{"label":"accuracy_score","shortLabel":"ACCUR","direction":"higher_better"},"macro_f1_score":{"label":"macro_f1_score","shortLabel":"MACRO","direction":"higher_better"},"weighted_f1_score":{"label":"weighted_f1_score","shortLabel":"WEIGH","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-geo-dnam-age-regression-lb-0010":{"meta":{"title":"LongevityBench (Peachy): GEO DNAm Age / Regression (LB-0010)","category":"Longevity > LongevityBench (Peachy)","description":"From a GEO DNA-methylation profile, the model estimates the donor's exact age in years.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":12.9556,"r2_score":0.4076,"spearmanr_score":0.5604,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","mean_absolute_error":17.0667,"r2_score":0.1895,"spearmanr_score":0.3415,"valid_fraction":1,"rank":5},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":18.7778,"r2_score":0.0255,"spearmanr_score":0.2235,"valid_fraction":1,"rank":6},{"model":"Grok 4.5, high","mean_absolute_error":14.933333333333334,"r2_score":0.43480901662458926,"spearmanr_score":0.7353894399990287,"valid_fraction":1,"rank":3},{"model":"GPT 5.6 Sol, high","mean_absolute_error":14.555555555555555,"r2_score":0.37590210464664187,"spearmanr_score":0.649678436777605,"valid_fraction":1,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":16.666666666666668,"r2_score":0.3278753317147536,"spearmanr_score":0.6535910161829801,"valid_fraction":0.8,"rank":4}],"name":"GEO DNAm Age / Regression (LB-0010)","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-nhanes-biomarkers-age-regression-lb-0038":{"meta":{"title":"LongevityBench (Peachy): NHANES Biomarkers Age / Regression (LB-0038)","category":"Longevity > LongevityBench (Peachy)","description":"From a person's NHANES clinical and blood-biomarker data, the model predicts their exact age in years.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"GPT 5.5, medium","mean_absolute_error":9.1333,"r2_score":0.6831,"spearmanr_score":0.8497,"valid_fraction":1,"rank":1},{"model":"Claude Opus 4.8, high","mean_absolute_error":10.7333,"r2_score":0.5051,"spearmanr_score":0.8104,"valid_fraction":1,"rank":6},{"model":"Gemini 3.1 Pro, high","mean_absolute_error":9.6,"r2_score":0.6236,"spearmanr_score":0.8213,"valid_fraction":1,"rank":2},{"model":"Grok 4.5, high","mean_absolute_error":10.088888888888889,"r2_score":0.637137343916321,"spearmanr_score":0.8437233156160314,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","mean_absolute_error":10.022222222222222,"r2_score":0.6097421976387558,"spearmanr_score":0.8324978377302357,"valid_fraction":1,"rank":4},{"model":"Claude Opus 5, high","mean_absolute_error":9.777777777777779,"r2_score":0.6447926742134451,"spearmanr_score":0.8230749789822117,"valid_fraction":1,"rank":3}],"name":"NHANES Biomarkers Age / Regression (LB-0038)","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"lon-longevitybench-peachy-gtex-transcriptomic-age-generation-lb-0026":{"meta":{"title":"LongevityBench (Peachy): GTEx Transcriptomic Age / Generation (LB-0026)","category":"Longevity > LongevityBench (Peachy)","description":"Built from GTEx. For a given tissue and age group, the model is shown 200 of the 400 most highly expressed genes and must generate the other 200 — recreating a real expression profile.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["jaccard_similarity","generation_f1","generation_recall","valid_fraction"],"leaderboard_metric":"jaccard_similarity","results":[{"model":"GPT 5.5, medium","jaccard_similarity":0.2328,"generation_f1":0.3707,"generation_recall":0.278,"valid_fraction":1,"rank":5},{"model":"Claude Opus 4.8, high","jaccard_similarity":0.2698,"generation_f1":0.4132,"generation_recall":0.3847,"valid_fraction":1,"rank":2},{"model":"Gemini 3.1 Pro, high","jaccard_similarity":0.2054,"generation_f1":0.3279,"generation_recall":0.2493,"valid_fraction":0.9333,"rank":6},{"model":"Grok 4.5, high","jaccard_similarity":0.26252144273728395,"generation_f1":0.4029609922611563,"generation_recall":0.3097777777777778,"valid_fraction":1,"rank":4},{"model":"GPT 5.6 Sol, high","jaccard_similarity":0.26945851107799235,"generation_f1":0.41942128081258523,"generation_recall":0.3145555555555556,"valid_fraction":1,"rank":3},{"model":"Claude Opus 5, high","jaccard_similarity":0.2836914564330038,"generation_f1":0.43767640575412003,"generation_recall":0.33,"valid_fraction":1,"rank":1}],"name":"GTEx Transcriptomic Age / Generation (LB-0026)","metric_info":{"jaccard_similarity":{"label":"jaccard_similarity","shortLabel":"JACCA","direction":"higher_better"},"generation_f1":{"label":"generation_f1","shortLabel":"GENER","direction":"higher_better"},"generation_recall":{"label":"generation_recall","shortLabel":"GENER","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"generation_f1","min":0,"max":1}},"lon-longevitybench-peachy-olink-proteomic-age-generation-lb-0058":{"meta":{"title":"LongevityBench (Peachy): Olink Proteomic Age / Generation (LB-0058)","category":"Longevity > LongevityBench (Peachy)","description":"Built from the Olink Explore 3072 panel (Kirsher et al., 2025). Shown 50 of the 100 most abundant plasma proteins for a person, the model must name the remaining 50.","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["jaccard_similarity","generation_f1","generation_recall","valid_fraction"],"leaderboard_metric":"jaccard_similarity","results":[{"model":"GPT 5.5, medium","jaccard_similarity":0.3915,"generation_f1":0.5293,"generation_recall":0.5292,"valid_fraction":1,"rank":2},{"model":"Claude Opus 4.8, high","jaccard_similarity":0.3688,"generation_f1":0.5071,"generation_recall":0.5071,"valid_fraction":1,"rank":4},{"model":"Gemini 3.1 Pro, high","jaccard_similarity":0.388,"generation_f1":0.5218,"generation_recall":0.5218,"valid_fraction":0.9808,"rank":3},{"model":"Grok 4.5, high","jaccard_similarity":0.3627485674274681,"generation_f1":0.4937370200434678,"generation_recall":0.4943589743589744,"valid_fraction":1,"rank":5},{"model":"GPT 5.6 Sol, high","jaccard_similarity":0.3985936607924107,"generation_f1":0.5372196322196323,"generation_recall":0.5371794871794873,"valid_fraction":1,"rank":1},{"model":"Claude Opus 5, high","jaccard_similarity":0.3486983590543092,"generation_f1":0.46923076923076923,"generation_recall":0.46923076923076923,"valid_fraction":0.8525641025641025,"rank":6}],"name":"Olink Proteomic Age / Generation (LB-0058)","metric_info":{"jaccard_similarity":{"label":"jaccard_similarity","shortLabel":"JACCA","direction":"higher_better"},"generation_f1":{"label":"generation_f1","shortLabel":"GENER","direction":"higher_better"},"generation_recall":{"label":"generation_recall","shortLabel":"GENER","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}},"insilico_score":{"metric":"generation_f1","min":0,"max":1}},"lon-longevitybench-peachy-synergyage-lifespan-regression-lb-0126":{"meta":{"title":"LongevityBench (Peachy): SynergyAge Lifespan / Regression (LB-0126)","category":"Longevity > LongevityBench (Peachy)","description":"synergyage_lifespan_regression","metric_types":[],"paper_link":"#","source_file":"longevity_peachy_ci.json","original_category":"Longevity & Aging"},"radar_keys":["mean_absolute_error","r2_score","spearmanr_score","valid_fraction"],"leaderboard_metric":"mean_absolute_error","results":[{"model":"Grok 4.5, high","mean_absolute_error":13.085999999999999,"r2_score":0.8035080444935069,"spearmanr_score":0.9030769230769231,"valid_fraction":0.9615384615384616,"rank":1},{"model":"GPT 5.6 Sol, high","mean_absolute_error":20.598000000000003,"r2_score":0.24758543874037264,"spearmanr_score":0.7923076923076923,"valid_fraction":0.9615384615384616,"rank":2},{"model":"Claude Opus 5, high","mean_absolute_error":23.6,"r2_score":0.11544562764770472,"spearmanr_score":0.39999999999999997,"valid_fraction":0.19230769230769232,"rank":3}],"name":"SynergyAge Lifespan / Regression (LB-0126)","metric_info":{"mean_absolute_error":{"label":"Mean absolute error","shortLabel":"MAE","direction":"lower_better"},"spearmanr_score":{"label":"Spearman correlation","shortLabel":"Spr","direction":"higher_better"},"valid_fraction":{"label":"Valid fraction","shortLabel":"Val","direction":"higher_better"}}},"str-genbio-hac-hic-suite-hac-retention-time":{"meta":{"title":"GenBio HAC HIC Suite: HAC Retention Time","category":"Therapeutic Biomolecules > GenBio HAC HIC Suite","description":"Evaluates heparin affinity chromatography (HAC) retention time, which measures antibody binding to heparin and reflects surface charge distribution and electrostatic interaction propensity. HAC is commonly used to assess nonspecific interaction liability and overall developability.","metric_types":[],"paper_link":"#","source_file":"genbio_hac_hic_ci.json","original_category":"Generative Biology"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.6490360569255783,"rank":1},{"model":"GPT 5.6-sol","spearman":0.5095130454305384,"rank":5},{"model":"Claude Opus 4.8, high","spearman":0.13800128206688456,"rank":9},{"model":"Opus 5.0","spearman":0.2712039453024409,"rank":7},{"model":"Gemini 3.1 Pro, high","spearman":0.5377044234812256,"rank":3},{"model":"Kimi K3","spearman":0.11473372484181285,"rank":10},{"model":"Grok 4.5, high","spearman":0.5191872106557298,"rank":4},{"model":"Claude Opus 5, high","spearman":0.2712039453024409,"rank":7},{"model":"GPT 5.6 Sol, high","spearman":0.5095130454305384,"rank":5},{"model":"Qwen 3.8 Max","spearman":0.5430841889016237,"rank":2}],"name":"HAC Retention Time","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-genbio-hac-hic-suite-hic-retention-time":{"meta":{"title":"GenBio HAC HIC Suite: HIC Retention Time","category":"Therapeutic Biomolecules > GenBio HAC HIC Suite","description":"Evaluates hydrophobic interaction chromatography (HIC) retention time, which measures interaction with a hydrophobic column and reflects exposed hydrophobic surface patches and aggregation risk. HIC is commonly used to assess nonspecific interaction liability and overall developability.","metric_types":[],"paper_link":"#","source_file":"genbio_hac_hic_ci.json","original_category":"Generative Biology"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.339331040576028,"rank":3},{"model":"GPT 5.6-sol","spearman":0.3597034282868534,"rank":1},{"model":"Claude Opus 4.8, high","spearman":0.10239581432613327,"rank":9},{"model":"Opus 5.0","spearman":0.16260945564637366,"rank":7},{"model":"Gemini 3.1 Pro, high","spearman":0.26685963451341477,"rank":5},{"model":"Kimi K3","spearman":0.08232136010576217,"rank":10},{"model":"Grok 4.5, high","spearman":0.18647129649532318,"rank":6},{"model":"Claude Opus 5, high","spearman":0.16260945564637366,"rank":7},{"model":"GPT 5.6 Sol, high","spearman":0.3597034282868534,"rank":1},{"model":"Qwen 3.8 Max","spearman":0.28527667876467394,"rank":4}],"name":"HIC Retention Time","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-polyreactivity-cho-polyreactivity":{"meta":{"title":"Polyreactivity: CHO Polyreactivity","category":"Therapeutic Biomolecules > Polyreactivity","description":"Evaluates antibody binding to CHO-cell membrane proteins (CHO) as a measure of polyreactivity. Polyreactivity is the propensity of an antibody to bind nonspecifically to unrelated antigens and is widely used as an indicator of nonspecific binding risk, which can impact antibody specificity, safety, and overall developability.","metric_types":[],"paper_link":"#","source_file":"genbio_polyreactivity_ci.json","original_category":"Biologics"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.13058390136813197,"rank":3},{"model":"GPT 5.6-sol","spearman":0.25864179723605024,"rank":1},{"model":"Claude Opus 4.8, high","spearman":0.06245934888216801,"rank":5},{"model":"Opus 5.0","spearman":0.026079652313482308,"rank":7},{"model":"Gemini 3.1 Pro, high","spearman":0.11431862263017559,"rank":4},{"model":"Kimi K3","spearman":0.025794232500286552,"rank":9},{"model":"Grok 4.5, high","spearman":0.046423672641765716,"rank":6},{"model":"Claude Opus 5, high","spearman":0.026079652313482308,"rank":7},{"model":"GPT 5.6 Sol, high","spearman":0.25864179723605024,"rank":1},{"model":"Qwen 3.8 Max","spearman":-0.009923250787909416,"rank":10}],"name":"CHO Polyreactivity","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-polyreactivity-ovalbumin-polyreactivity":{"meta":{"title":"Polyreactivity: Ovalbumin Polyreactivity","category":"Therapeutic Biomolecules > Polyreactivity","description":"Evaluates antibody binding to Ovalbumin as a measure of polyreactivity. Polyreactivity is the propensity of an antibody to bind nonspecifically to unrelated antigens and is widely used as an indicator of nonspecific binding risk, which can impact antibody specificity, safety, and overall developability.","metric_types":[],"paper_link":"#","source_file":"genbio_polyreactivity_ci.json","original_category":"Biologics"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.15180702492015258,"rank":3},{"model":"GPT 5.6-sol","spearman":0.29338085306977457,"rank":1},{"model":"Claude Opus 4.8, high","spearman":0.05029673634881512,"rank":5},{"model":"Opus 5.0","spearman":-0.0020184966688742956,"rank":6},{"model":"Gemini 3.1 Pro, high","spearman":0.11387730253239689,"rank":4},{"model":"Kimi K3","spearman":-0.05911504975210121,"rank":10},{"model":"Grok 4.5, high","spearman":-0.0024982412941896496,"rank":8},{"model":"Claude Opus 5, high","spearman":-0.0020184966688742956,"rank":6},{"model":"GPT 5.6 Sol, high","spearman":0.29338085306977457,"rank":1},{"model":"Qwen 3.8 Max","spearman":-0.025416092548976452,"rank":9}],"name":"Ovalbumin Polyreactivity","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-sec-smac-sec-monomer":{"meta":{"title":"SEC SMAC: SEC %Monomer","category":"Therapeutic Biomolecules > SEC SMAC","description":"SEC %Monomer, assessed by size-exclusion chromatography, reflects the proportion of monomeric antibody and serves as an indicator of molecular integrity and aggregation risk, impacting stability, efficacy, manufacturability, and overall developability.","metric_types":[],"paper_link":"#","source_file":"genbio_sec_smac_ci.json","original_category":"Biologics"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.020878393256096466,"rank":7},{"model":"GPT 5.6-sol","spearman":-0.006153737957946655,"rank":9},{"model":"Claude Opus 4.8, high","spearman":0.07801008648050708,"rank":4},{"model":"Opus 5.0","spearman":0.30963787221080713,"rank":1},{"model":"Gemini 3.1 Pro, high","spearman":0.04683579611727244,"rank":6},{"model":"Kimi K3","spearman":0.08716953930514863,"rank":3},{"model":"Grok 4.5, high","spearman":0.07558916988575569,"rank":5},{"model":"Claude Opus 5, high","spearman":0.30963787221080713,"rank":1},{"model":"GPT 5.6 Sol, high","spearman":-0.006153737957946655,"rank":9},{"model":"Qwen 3.8 Max","spearman":-0.005357059150484054,"rank":8}],"name":"SEC %Monomer","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-sec-smac-smac":{"meta":{"title":"SEC SMAC: SMAC","category":"Therapeutic Biomolecules > SEC SMAC","description":"SMAC measures antibody retention time using standup monolayer affinity chromatography, where longer retention reflects stronger surface interactions and provides insight into surface interaction propensity and developability-related biophysical characteristics.","metric_types":[],"paper_link":"#","source_file":"genbio_sec_smac_ci.json","original_category":"Biologics"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.22468164584463604,"rank":5},{"model":"GPT 5.6-sol","spearman":0.29568896253735705,"rank":2},{"model":"Claude Opus 4.8, high","spearman":0.07457663124470228,"rank":10},{"model":"Opus 5.0","spearman":0.14983993802549417,"rank":7},{"model":"Gemini 3.1 Pro, high","spearman":0.10367168128563183,"rank":9},{"model":"Kimi K3","spearman":0.240495597908878,"rank":4},{"model":"Grok 4.5, high","spearman":0.1942907596028653,"rank":6},{"model":"Claude Opus 5, high","spearman":0.14983993802549417,"rank":7},{"model":"GPT 5.6 Sol, high","spearman":0.29568896253735705,"rank":2},{"model":"Qwen 3.8 Max","spearman":0.3470989716492995,"rank":1}],"name":"SMAC","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-tm-and-titer-tm":{"meta":{"title":"Tm and Titer: Tm","category":"Therapeutic Biomolecules > Tm and Titer","description":"Evaluates Tm using Spearman correlation.","metric_types":[],"paper_link":"#","source_file":"genbio_tm_titer_ci.json","original_category":"Biologics"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":0.27708782426069234,"rank":2},{"model":"GPT 5.6-sol","spearman":0.2416441428965922,"rank":5},{"model":"Claude Opus 4.8, high","spearman":0.365578578523314,"rank":1},{"model":"Opus 5.0","spearman":0.25481134385233367,"rank":3},{"model":"Gemini 3.1 Pro, high","spearman":0.07488570032988526,"rank":8},{"model":"Kimi K3","spearman":0.048399398721983365,"rank":9},{"model":"Grok 4.5, high","spearman":0.03690122060630206,"rank":10},{"model":"Claude Opus 5, high","spearman":0.25481134385233367,"rank":3},{"model":"GPT 5.6 Sol, high","spearman":0.2416441428965922,"rank":5},{"model":"Qwen 3.8 Max","spearman":0.08637696637920954,"rank":7}],"name":"Tm","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}},"str-tm-and-titer-titer":{"meta":{"title":"Tm and Titer: Titer","category":"Therapeutic Biomolecules > Tm and Titer","description":"Evaluates titer using Spearman correlation.","metric_types":[],"paper_link":"#","source_file":"genbio_tm_titer_ci.json","original_category":"Biologics"},"radar_keys":["spearman"],"leaderboard_metric":"spearman","results":[{"model":"GPT 5.5, medium","spearman":-0.09847690153064981,"rank":8},{"model":"GPT 5.6-sol","spearman":-0.06482843499135031,"rank":5},{"model":"Claude Opus 4.8, high","spearman":-0.050258441851106235,"rank":4},{"model":"Opus 5.0","spearman":0.12342857190553248,"rank":1},{"model":"Gemini 3.1 Pro, high","spearman":0.0779725921008153,"rank":3},{"model":"Kimi K3","spearman":-0.07857888618420666,"rank":7},{"model":"Grok 4.5, high","spearman":-0.10704195844606826,"rank":9},{"model":"Claude Opus 5, high","spearman":0.12342857190553248,"rank":1},{"model":"GPT 5.6 Sol, high","spearman":-0.06482843499135031,"rank":5},{"model":"Qwen 3.8 Max","spearman":-0.15789967182372247,"rank":10}],"name":"Titer","metric_info":{"spearman":{"label":"Spearman's Rank Correlation","shortLabel":"SPEAR","direction":"higher_better"}}}},"model_scope":["GPT 5.5, medium","GPT 5.6-sol","Claude Opus 4.8, high","Opus 5.0","Gemini 3.1 Pro, high","Kimi K3","Grok 4.5, high","Claude Opus 5, high","GPT 5.6 Sol, high","GPT 5.6 Sol, medium","GPT 5.6 Terra, high","Kimi K3, max","GPT 5.6 Luna, high","GPT 5.6 Terra, medium","GPT 5.6 Luna, medium","RetroChimera","RetroChimera-MCTS","Aizynthfinder-MCTS","Qwen 3.8 Max"]},"phases":{"definitions":[{"id":"disease-modelling","name":"Disease Modelling","order":0},{"id":"target-discovery","name":"Target Discovery","order":1},{"id":"target-to-hit","name":"Target-to-Hit","order":2},{"id":"hit-to-lead","name":"Hit-to-Lead","order":3},{"id":"lead-optimization","name":"Lead Optimization","order":4},{"id":"preclinical-candidate","name":"Preclinical Candidate","order":5},{"id":"ind-enabling","name":"IND-Enabling","order":6},{"id":"phase-1","name":"Phase 1","order":7},{"id":"phase-2","name":"Phase 2","order":8},{"id":"phase-3","name":"Phase 3","order":9},{"id":"phase-4","name":"Phase 4","order":10},{"id":"all-phases","name":"All phases","order":11}],"assignments":{"dis-targetbench-by-disease-group-target-prediction-for-cancer-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-cardiovascular-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-fibrotic-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-inflammatory-immunological-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-mental-behavioral-disorders":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-metabolic-endocrine-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-neurological-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-ophthalmic-diseases":"target-discovery","dis-targetbench-by-disease-group-target-prediction-for-reproductive-health-conditions":"target-discovery","med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-all-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-fragment-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-only-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-crossdocked-pocket-pharmacophore-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-plinder-all-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-fragment-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-only-condition":"target-to-hit","med-bench3dfit-molecular-generation-with-3d-conditions-plinder-pocket-pharmacophore-condition":"target-to-hit","med-fgbench-functional-groups-reasoning-functional-group-comparison-direction":"lead-optimization","med-fgbench-functional-groups-reasoning-functional-group-comparison-value":"lead-optimization","med-fgbench-functional-groups-reasoning-functional-group-interaction-direction":"lead-optimization","med-fgbench-functional-groups-reasoning-functional-group-interaction-value":"lead-optimization","med-fgbench-functional-groups-reasoning-single-functional-group-direction":"lead-optimization","med-fgbench-functional-groups-reasoning-single-functional-group-value":"lead-optimization","med-gpcr-affinity-suite-5-ht1a-ic50":"target-to-hit","med-gpcr-affinity-suite-5-ht1a-ki":"target-to-hit","med-gpcr-affinity-suite-5-ht1c-ki":"target-to-hit","med-gpcr-affinity-suite-5-ht2a-ic50":"target-to-hit","med-gpcr-affinity-suite-5-ht2a-ki":"target-to-hit","med-gpcr-affinity-suite-5-ht2b-ki":"target-to-hit","med-gpcr-affinity-suite-5-ht5a-ki":"target-to-hit","med-gpcr-affinity-suite-5-ht6-ic50":"target-to-hit","med-gpcr-affinity-suite-5-ht6-ki":"target-to-hit","med-gpcr-affinity-suite-5-ht7-ki":"target-to-hit","med-gpcr-affinity-suite-a1-ic50":"target-to-hit","med-gpcr-affinity-suite-a1-ki":"target-to-hit","med-gpcr-affinity-suite-a2a-ic50":"target-to-hit","med-gpcr-affinity-suite-a2a-ki":"target-to-hit","med-gpcr-affinity-suite-a2b-ic50":"target-to-hit","med-gpcr-affinity-suite-a2b-ki":"target-to-hit","med-gpcr-affinity-suite-a3-ki":"target-to-hit","med-gpcr-affinity-suite-alpha1a-ki":"target-to-hit","med-gpcr-affinity-suite-b1-ic50":"target-to-hit","med-gpcr-affinity-suite-b1-ki":"target-to-hit","med-gpcr-affinity-suite-casr-ic50":"target-to-hit","med-gpcr-affinity-suite-cb1-ic50":"target-to-hit","med-gpcr-affinity-suite-cb1-ki":"target-to-hit","med-gpcr-affinity-suite-cb2-ic50":"target-to-hit","med-gpcr-affinity-suite-cb2-ki":"target-to-hit","med-gpcr-affinity-suite-ccr1-ic50":"target-to-hit","med-gpcr-affinity-suite-ccr2-ic50":"target-to-hit","med-gpcr-affinity-suite-ccr3-ic50":"target-to-hit","med-gpcr-affinity-suite-ccr5-ic50":"target-to-hit","med-gpcr-affinity-suite-crhr1-ic50":"target-to-hit","med-gpcr-affinity-suite-crhr1-ki":"target-to-hit","med-gpcr-affinity-suite-cxcr3-ic50":"target-to-hit","med-gpcr-affinity-suite-cxcr4-ic50":"target-to-hit","med-gpcr-affinity-suite-d1-ki":"target-to-hit","med-gpcr-affinity-suite-d2-ic50":"target-to-hit","med-gpcr-affinity-suite-d2-ki":"target-to-hit","med-gpcr-affinity-suite-d3-ki":"target-to-hit","med-gpcr-affinity-suite-d4-ki":"target-to-hit","med-gpcr-affinity-suite-ep2-ic50":"target-to-hit","med-gpcr-affinity-suite-ep4-ic50":"target-to-hit","med-gpcr-affinity-suite-eta-ic50":"target-to-hit","med-gpcr-affinity-suite-ghs-r-ic50":"target-to-hit","med-gpcr-affinity-suite-glucagon-ic50":"target-to-hit","med-gpcr-affinity-suite-glucagon-ki":"target-to-hit","med-gpcr-affinity-suite-gnrhr-ic50":"target-to-hit","med-gpcr-affinity-suite-gnrhr-ki":"target-to-hit","med-gpcr-affinity-suite-gpr44-ic50":"target-to-hit","med-gpcr-affinity-suite-gpr44-ki":"target-to-hit","med-gpcr-affinity-suite-il8rb-ic50":"target-to-hit","med-gpcr-affinity-suite-machrm1-ic50":"target-to-hit","med-gpcr-affinity-suite-machrm1-ki":"target-to-hit","med-gpcr-affinity-suite-machrm2-ki":"target-to-hit","med-gpcr-affinity-suite-machrm3-ki":"target-to-hit","med-gpcr-affinity-suite-machrm4-ic50":"target-to-hit","med-gpcr-affinity-suite-machrm4-ki":"target-to-hit","med-gpcr-affinity-suite-mc3r-ki":"target-to-hit","med-gpcr-affinity-suite-mc4r-ic50":"target-to-hit","med-gpcr-affinity-suite-mc4r-ki":"target-to-hit","med-gpcr-affinity-suite-mc5r-ki":"target-to-hit","med-gpcr-affinity-suite-mglur1-ic50":"target-to-hit","med-gpcr-affinity-suite-mglur2-ic50":"target-to-hit","med-gpcr-affinity-suite-mglur5-ic50":"target-to-hit","med-gpcr-affinity-suite-mglur5-ki":"target-to-hit","med-gpcr-affinity-suite-nk1-ki":"target-to-hit","med-gpcr-affinity-suite-nk2-ic50":"target-to-hit","med-gpcr-affinity-suite-nk2-ki":"target-to-hit","med-gpcr-affinity-suite-nk3-ic50":"target-to-hit","med-gpcr-affinity-suite-nk3-ki":"target-to-hit","med-gpcr-affinity-suite-npy1r-ki":"target-to-hit","med-gpcr-affinity-suite-npy2r-ic50":"target-to-hit","med-gpcr-affinity-suite-npy5r-ic50":"target-to-hit","med-gpcr-affinity-suite-npy5r-ki":"target-to-hit","med-gpcr-affinity-suite-oprd1-ic50":"target-to-hit","med-gpcr-affinity-suite-oprd1-ki":"target-to-hit","med-gpcr-affinity-suite-oprk1-ic50":"target-to-hit","med-gpcr-affinity-suite-oprk1-ki":"target-to-hit","med-gpcr-affinity-suite-oprm1-ic50":"target-to-hit","med-gpcr-affinity-suite-oprm1-ki":"target-to-hit","med-gpcr-affinity-suite-ox1-ic50":"target-to-hit","med-gpcr-affinity-suite-ox1-ki":"target-to-hit","med-gpcr-affinity-suite-ox2-ic50":"target-to-hit","med-gpcr-affinity-suite-ox2-ki":"target-to-hit","med-gpcr-affinity-suite-thrombin-ic50":"target-to-hit","med-gpcr-affinity-suite-v1a-ic50":"target-to-hit","med-gpcr-affinity-suite-v1a-ki":"target-to-hit","med-gpcr-affinity-suite-v2-ic50":"target-to-hit","med-gpcr-affinity-suite-v2-ki":"target-to-hit","med-medchem-holdout-ic50-prediction-internal-holdout":"target-to-hit","med-ism-benchmarks-admet-aggregation-potential":"lead-optimization","med-ism-benchmarks-admet-ames":"lead-optimization","med-ism-benchmarks-admet-bcrp-inhibition":"lead-optimization","med-ism-benchmarks-admet-blood-brain-barrier-penetration":"lead-optimization","med-ism-benchmarks-admet-caco-2-efflux-ratio":"lead-optimization","med-ism-benchmarks-admet-caco-2-permeability":"lead-optimization","med-ism-benchmarks-admet-clearance-human-hepatocytes":"lead-optimization","med-ism-benchmarks-admet-clearance-human-liver-microsomes":"lead-optimization","med-ism-benchmarks-admet-clearance-mouse-liver-microsomes":"lead-optimization","med-ism-benchmarks-admet-clearance-rat-liver-microsomes":"lead-optimization","med-ism-benchmarks-admet-cyp1a2-inhibition-ic50":"lead-optimization","med-ism-benchmarks-admet-cyp2c19-inhibition-ic50":"lead-optimization","med-ism-benchmarks-admet-cyp2c9-inhibition-ic50":"lead-optimization","med-ism-benchmarks-admet-cyp2d6-inhibition-ic50":"lead-optimization","med-ism-benchmarks-admet-cyp3a4-inhibition-ic50":"lead-optimization","med-ism-benchmarks-admet-half-life-human":"lead-optimization","med-ism-benchmarks-admet-hek293-cc50":"lead-optimization","med-ism-benchmarks-admet-hepg2-cc50":"lead-optimization","med-ism-benchmarks-admet-herg-ic50":"lead-optimization","med-ism-benchmarks-admet-kinetic-solubility":"lead-optimization","med-ism-benchmarks-admet-logd":"lead-optimization","med-ism-benchmarks-admet-logp":"lead-optimization","med-ism-benchmarks-admet-mdck-mdr1-efflux-ratio":"lead-optimization","med-ism-benchmarks-admet-mdck-mdr1-permeability":"lead-optimization","med-ism-benchmarks-admet-oatp1b1-inhibition":"lead-optimization","med-ism-benchmarks-admet-p-gp-inhibition":"lead-optimization","med-ism-benchmarks-admet-pampa-permeability":"lead-optimization","med-ism-benchmarks-admet-ppb-human":"lead-optimization","med-ism-benchmarks-admet-ppb-mouse":"lead-optimization","med-ism-benchmarks-admet-ppb-rat":"lead-optimization","med-ism-benchmarks-admet-solubility-in-dmso":"lead-optimization","med-ism-benchmarks-admet-thermodynamic-solubility":"lead-optimization","med-ism-benchmarks-admet-vdss":"lead-optimization","med-ism-benchmarks-protacs-pk-bioavailability-f-upon-oral-administration-in-mice":"lead-optimization","med-ism-benchmarks-protacs-pk-dnauc-dose-normalized-auc-0-t-ngmlhmg-upon-iv-administration-in-mice":"lead-optimization","med-ism-benchmarks-protacs-pk-half-life-t12-h-upon-iv-administration-in-mice":"lead-optimization","med-molecular-mechanics-energy-conformation-ranking":"target-to-hit","med-molecular-mechanics-energy-mmff-energy-prediction":"target-to-hit","med-mumo-multiple-property-optimization-bbbp-drd2-plogp":"lead-optimization","med-mumo-multiple-property-optimization-bbbp-drd2-plogp-qed":"lead-optimization","med-mumo-multiple-property-optimization-bbbp-drd2-qed":"lead-optimization","med-mumo-multiple-property-optimization-bbbp-plogp-qed":"lead-optimization","med-mumo-multiple-property-optimization-drd2-plogp-qed":"lead-optimization","med-pdbbind-restore-interactions-3d-interactions-restoration":"target-to-hit","med-pkis2-kinase-inhibition-suite-pkis2-egfr":"target-to-hit","med-pkis2-kinase-inhibition-suite-pkis2-kit":"target-to-hit","med-pkis2-kinase-inhibition-suite-pkis2-ret":"target-to-hit","med-tdc-admet-ames-mutagenicity":"lead-optimization","med-tdc-admet-bbb-penetration":"lead-optimization","med-tdc-admet-bioavailability":"lead-optimization","med-tdc-admet-caco-2-permeability":"lead-optimization","med-tdc-admet-clearance-hepatocyte":"lead-optimization","med-tdc-admet-clearance-microsome":"lead-optimization","med-tdc-admet-cyp2c9-inhibition":"lead-optimization","med-tdc-admet-cyp2c9-substrate":"lead-optimization","med-tdc-admet-cyp2d6-inhibition":"lead-optimization","med-tdc-admet-cyp2d6-substrate":"lead-optimization","med-tdc-admet-cyp3a4-inhibition":"lead-optimization","med-tdc-admet-cyp3a4-substrate":"lead-optimization","med-tdc-admet-dili":"lead-optimization","med-tdc-admet-half-life":"lead-optimization","med-tdc-admet-herg-blockers":"lead-optimization","med-tdc-admet-hia":"lead-optimization","med-tdc-admet-ld50-acute-toxicity":"lead-optimization","med-tdc-admet-lipophilicity":"lead-optimization","med-tdc-admet-p-gp-inhibition":"lead-optimization","med-tdc-admet-ppbr":"lead-optimization","med-tdc-admet-solubility":"lead-optimization","med-tdc-admet-vdss":"lead-optimization","che-reaction-plausibility-reaction-plausibility":"lead-optimization","che-retrosynthesis-suite-single-step-retrosynthesis-ursa-expert-2026-dataset":"lead-optimization","che-retrosynthesis-suite-single-step-retrosynthesis-uspto-50k-test-sample-dataset":"lead-optimization","syn-retrosynthesis-suite-single-step-retrosynthesis-top-15-task-ursa-expert-2026-dataset":"lead-optimization","syn-retrosynthesis-suite-single-step-retrosynthesis-top-15-task-uspto-50k-test-sample-dataset":"lead-optimization","syn-msrs-multi-step-retrosynthesis-ursa-drugsclinicals-2026-test-set":"lead-optimization","syn-msrs-multi-step-retrosynthesis-ursa-expert-2026-test-set":"lead-optimization","cli-clinbench-quarterly-all-phases":"all-phases","cli-clinbench-quarterly-all-phases-2025q1":"all-phases","cli-clinbench-quarterly-all-phases-2025q2":"all-phases","cli-clinbench-quarterly-all-phases-2025q3":"all-phases","cli-clinbench-quarterly-all-phases-2025q4":"all-phases","cli-clinbench-quarterly-all-phases-2026q1":"all-phases","cli-clinbench-quarterly-phase-1":"all-phases","cli-clinbench-quarterly-phase-1-2025q1":"all-phases","cli-clinbench-quarterly-phase-1-2025q2":"all-phases","cli-clinbench-quarterly-phase-1-2025q3":"all-phases","cli-clinbench-quarterly-phase-1-2025q4":"all-phases","cli-clinbench-quarterly-phase-1-2026q1":"all-phases","cli-clinbench-quarterly-phase-2":"all-phases","cli-clinbench-quarterly-phase-2-2025q1":"all-phases","cli-clinbench-quarterly-phase-2-2025q2":"all-phases","cli-clinbench-quarterly-phase-2-2025q3":"all-phases","cli-clinbench-quarterly-phase-2-2025q4":"all-phases","cli-clinbench-quarterly-phase-2-2026q1":"all-phases","cli-clinbench-quarterly-phase-3":"all-phases","cli-clinbench-quarterly-phase-3-2025q1":"all-phases","cli-clinbench-quarterly-phase-3-2025q2":"all-phases","cli-clinbench-quarterly-phase-3-2025q3":"all-phases","cli-clinbench-quarterly-phase-3-2025q4":"all-phases","cli-clinbench-quarterly-phase-3-2026q1":"all-phases","cli-clinbench-quarterly-phase-4":"all-phases","cli-clinbench-quarterly-phase-4-2025q1":"all-phases","cli-clinbench-quarterly-phase-4-2025q2":"all-phases","cli-clinbench-quarterly-phase-4-2025q3":"all-phases","cli-clinbench-quarterly-phase-4-2025q4":"all-phases","cli-clinbench-quarterly-phase-4-2026q1":"all-phases","lon-longevitybench-peachy-geo-dnam-age-multiclass-lb-0002":"all-phases","lon-longevitybench-peachy-geo-dnam-age-pairwise-lb-0006":"all-phases","lon-longevitybench-peachy-geo-dnam-age-regression-lb-0010":"all-phases","lon-longevitybench-peachy-gtex-transcriptomic-age-generation-lb-0026":"all-phases","lon-longevitybench-peachy-gtex-transcriptomic-age-multiclass-lb-0014":"all-phases","lon-longevitybench-peachy-gtex-transcriptomic-age-pairwise-lb-0018":"all-phases","lon-longevitybench-peachy-nhanes-biomarkers-age-multiclass-lb-0030":"all-phases","lon-longevitybench-peachy-nhanes-biomarkers-age-pairwise-lb-0034":"all-phases","lon-longevitybench-peachy-nhanes-biomarkers-age-regression-lb-0038":"all-phases","lon-longevitybench-peachy-nhanes-biomarkers-mortality-binary-lb-0042":"all-phases","lon-longevitybench-peachy-nhanes-biomarkers-mortality-multiclass-lb-0046":"all-phases","lon-longevitybench-peachy-nhanes-biomarkers-mortality-pairwise-lb-0050":"all-phases","lon-longevitybench-peachy-olink-proteomic-age-generation-lb-0058":"all-phases","lon-longevitybench-peachy-olink-proteomic-age-pairwise-lb-0066":"all-phases","lon-longevitybench-peachy-opengenes-expression-binary_masked-lb-0086":"all-phases","lon-longevitybench-peachy-synergyage-lifespan-pairwise-ablated-lb-0122":"all-phases","lon-longevitybench-peachy-synergyage-lifespan-regression-lb-0126":"all-phases","lon-longevitybench-peachy-tcga-cancer-survival-pairwise-lb-0134":"all-phases","str-genbio-hac-hic-suite-hac-retention-time":"lead-optimization","str-genbio-hac-hic-suite-hic-retention-time":"lead-optimization","str-polyreactivity-cho-polyreactivity":"lead-optimization","str-polyreactivity-ovalbumin-polyreactivity":"lead-optimization","str-sec-smac-sec-monomer":"lead-optimization","str-sec-smac-smac":"lead-optimization","str-tm-and-titer-titer":"lead-optimization","str-tm-and-titer-tm":"lead-optimization","med-kinases-affinity-suite-abl1":"target-to-hit","med-kinases-affinity-suite-akt1":"target-to-hit","med-kinases-affinity-suite-akt2":"target-to-hit","med-kinases-affinity-suite-atm":"target-to-hit","med-kinases-affinity-suite-atr":"target-to-hit","med-kinases-affinity-suite-aurka":"target-to-hit","med-kinases-affinity-suite-aurkb":"target-to-hit","med-kinases-affinity-suite-axl":"target-to-hit","med-kinases-affinity-suite-braf":"target-to-hit","med-kinases-affinity-suite-btk":"target-to-hit","med-kinases-affinity-suite-bub1":"target-to-hit","med-kinases-affinity-suite-c-met":"target-to-hit","med-kinases-affinity-suite-c-src":"target-to-hit","med-kinases-affinity-suite-cdk2":"target-to-hit","med-kinases-affinity-suite-cdk7":"target-to-hit","med-kinases-affinity-suite-cdk9":"target-to-hit","med-kinases-affinity-suite-chek1":"target-to-hit","med-kinases-affinity-suite-csnk1d":"target-to-hit","med-kinases-affinity-suite-csf1r":"target-to-hit","med-kinases-affinity-suite-ddr1":"target-to-hit","med-kinases-affinity-suite-egrf":"target-to-hit","med-kinases-affinity-suite-erbb2":"target-to-hit","med-kinases-affinity-suite-erbb4":"target-to-hit","med-kinases-affinity-suite-fgfr2":"target-to-hit","med-kinases-affinity-suite-fgfr3":"target-to-hit","med-kinases-affinity-suite-flt3":"target-to-hit","med-kinases-affinity-suite-gsk3b":"target-to-hit","med-kinases-affinity-suite-igf1r":"target-to-hit","med-kinases-affinity-suite-ikbkb":"target-to-hit","med-kinases-affinity-suite-insr":"target-to-hit","med-kinases-affinity-suite-irak4":"target-to-hit","med-kinases-affinity-suite-jak1":"target-to-hit","med-kinases-affinity-suite-jak2":"target-to-hit","med-kinases-affinity-suite-jak3":"target-to-hit","med-kinases-affinity-suite-kit":"target-to-hit","med-kinases-affinity-suite-lck":"target-to-hit","med-kinases-affinity-suite-map2k1":"target-to-hit","med-kinases-affinity-suite-map4k4":"target-to-hit","med-kinases-affinity-suite-mapk10":"target-to-hit","med-kinases-affinity-suite-mapk11":"target-to-hit","med-kinases-affinity-suite-mapk14":"target-to-hit","med-kinases-affinity-suite-mapk8":"target-to-hit","med-kinases-affinity-suite-mapk9":"target-to-hit","med-kinases-affinity-suite-mapkapk2":"target-to-hit","med-kinases-affinity-suite-mknk":"target-to-hit","med-kinases-affinity-suite-mknk2":"target-to-hit","med-kinases-affinity-suite-ntrk-1":"target-to-hit","med-kinases-affinity-suite-pdgfr":"target-to-hit","med-kinases-affinity-suite-pdpk1":"target-to-hit","med-kinases-affinity-suite-pik3ca":"target-to-hit","med-kinases-affinity-suite-pik3cg":"target-to-hit","med-kinases-affinity-suite-plk1":"target-to-hit","med-kinases-affinity-suite-plk2":"target-to-hit","med-kinases-affinity-suite-plk4":"target-to-hit","med-kinases-affinity-suite-prkca":"target-to-hit","med-kinases-affinity-suite-prkdc":"target-to-hit","med-kinases-affinity-suite-ptk2":"target-to-hit","med-kinases-affinity-suite-ret":"target-to-hit","med-kinases-affinity-suite-rock2":"target-to-hit","med-kinases-affinity-suite-rsk2":"target-to-hit","med-kinases-affinity-suite-tek":"target-to-hit","med-kinases-affinity-suite-tgfbr1":"target-to-hit","med-kinases-affinity-suite-tnik":"target-to-hit","med-kinases-affinity-suite-ttk":"target-to-hit","med-kinases-affinity-suite-tyk2":"target-to-hit","med-kinases-affinity-suite-vegfr2":"target-to-hit","med-kinases-affinity-suite-zap70":"target-to-hit"}},"leaderboardHiddenModels":["GPT 5.6-sol","Kimi K3","Opus 5.0","GPT 5.6 Sol, medium","GPT 5.6 Terra, high","Kimi K3, max","GPT 5.6 Luna, high","GPT 5.6 Terra, medium","GPT 5.6 Luna, medium","RetroChimera","RetroChimera-MCTS","Aizynthfinder-MCTS","LFM2-MMAI-Chem-SSRS","Qwen 3.8 Max"],"insilicoScoreConfig":{"benchmarks":{}},"tabVisibility":{"leaderboard":true,"explorer":true,"mmai":false},"meta":{"totalBenchmarks":310,"contributionFiles":6,"errors":0}}