{"id":"W2903425689","doi":"10.3389/fphar.2020.565644","title":"Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models","year":2020,"lang":"en","type":"preprint","venue":"Frontiers in Pharmacology","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":84,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto; Canadian Institute for Advanced Research","funders":"","keywords":"Benchmarking; Generative grammar; Computer science; Set (abstract data type); Machine learning; Generative model; Code (set theory); Training set; Quality (philosophy); Chemical space; Artificial intelligence; Data mining; Bioinformatics; Programming language; Biology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001321634,0.0005548368,0.0008917885,0.0003355311,0.0001598922,0.000227187,0.001371335,0.000531829,0.0002916779],"category_scores_gemma":[0.0001301219,0.0006197084,0.0001944671,0.0002308524,0.0002160027,0.0002778092,0.001147741,0.0007391019,0.00001990178],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003871379,"about_ca_system_score_gemma":0.0003982398,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006584785,"about_ca_topic_score_gemma":0.00000430574,"domain_scores_codex":[0.9958427,0.0004302557,0.000834824,0.001552813,0.0004726139,0.0008667858],"domain_scores_gemma":[0.998583,0.0000721094,0.0005531391,0.0004555402,0.0001310003,0.0002051953],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00009887996,0.0000346789,0.00003725555,0.0001599153,0.00002402365,0.00008464236,0.0002976926,0.422266,0.5650601,0.0004083744,0.01076421,0.000764281],"study_design_scores_gemma":[0.0007153991,0.00009440693,0.000008619561,0.0000394034,0.00008325262,0.000006747289,0.00001710616,0.7061049,0.2728706,0.01878858,0.0008060054,0.0004650063],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4346561,0.0005860588,0.5470603,0.001005625,0.01473649,0.001432818,0.0001325195,0.0001426274,0.000247481],"genre_scores_gemma":[0.6544709,0.00006878812,0.3416612,0.001876789,0.0006311695,0.0009123819,0.0002793958,0.00008568399,0.00001368055],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.2921895,"threshold_uncertainty_score":0.9996254,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03716816222312058,"score_gpt":0.3199947715576369,"score_spread":0.2828266093345164,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}