{"id":"W4391431291","doi":"10.1021/acs.jcim.3c01774","title":"Real-World Molecular Out-Of-Distribution: Specification and Investigation","year":2024,"lang":"en","type":"article","venue":"Journal of Chemical Information and Modeling","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":33,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"Mitacs","keywords":"Benchmark (surveying); Computer science; Calibration; Generalization; Protocol (science); Software deployment; Set (abstract data type); Sample (material); Covariate; Mood; Data mining; Artificial intelligence; Machine learning; Statistics; Mathematics; Software engineering; Psychology","routes":{"ca_aff":true,"ca_fund":true,"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":[],"consensus_categories":[],"category_scores_codex":[0.000492333,0.00005454646,0.00009615356,0.0001294851,0.0000175624,0.0001555572,0.00009356452,0.0000285917,0.000001030694],"category_scores_gemma":[0.00007140004,0.00004847816,0.00003453335,0.0001828101,0.0000253465,0.001770898,0.0000468839,0.0001120988,9.242394e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003881449,"about_ca_system_score_gemma":0.00008686329,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000002536611,"about_ca_topic_score_gemma":8.887712e-8,"domain_scores_codex":[0.9991656,0.00002337304,0.0004829982,0.00005617505,0.0002182405,0.00005364467],"domain_scores_gemma":[0.9994323,0.00007755329,0.0001545961,0.00006008178,0.0002112968,0.00006423135],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003175894,0.00001934333,0.00004536804,0.0003333491,0.00005467093,0.000004099885,0.003874693,0.1317302,0.0535903,0.6430505,0.0005210948,0.1667446],"study_design_scores_gemma":[0.00009826601,0.00001153122,0.00002859138,0.00008851493,0.000007680383,0.00001935395,0.00002486512,0.9535031,0.017715,0.02777862,0.0006762384,0.00004827998],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.243871,0.0001558554,0.7547497,0.0008500008,0.0001264845,0.00002779911,0.000001827751,0.00001257116,0.0002047091],"genre_scores_gemma":[0.9419727,0.0000978369,0.05779959,0.00007009714,0.00004314317,6.513326e-7,0.00001205135,0.000001754489,0.000002155041],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8217728,"threshold_uncertainty_score":0.1976882,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04529205064243014,"score_gpt":0.3197108155328022,"score_spread":0.274418764890372,"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."}}