{"id":"W4385402127","doi":"10.26434/chemrxiv-2023-q11q4-v2","title":"Real-World Molecular Out-Of-Distribution: Specification and Investigation","year":2023,"lang":"en","type":"preprint","venue":"ChemRxiv","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"Mitacs","keywords":"Computer science; Software deployment; Generalization; Protocol (science); Calibration; Benchmark (surveying); Mood; Code (set theory); Sample (material); Set (abstract data type); Artificial intelligence; Data mining; Machine learning; Software engineering; Statistics; Programming language; Mathematics; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02279486,0.0009953139,0.001076417,0.0007245029,0.0007402559,0.002463541,0.003076718,0.002234963,0.001767002],"category_scores_gemma":[0.1015642,0.0005757822,0.001214633,0.0007988711,0.002743401,0.003467049,0.003507649,0.003665759,0.0003177153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002231522,"about_ca_system_score_gemma":0.001632818,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001782368,"about_ca_topic_score_gemma":0.001291663,"domain_scores_codex":[0.990254,0.00549631,0.000374956,0.001650111,0.001705077,0.0005196308],"domain_scores_gemma":[0.9308568,0.05414385,0.003934798,0.007617164,0.002741625,0.0007057518],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006780426,0.0005051448,0.01780897,0.0006824712,0.000207095,0.0006095296,0.0004690522,0.7858044,0.006605559,0.1275165,0.007977915,0.05113522],"study_design_scores_gemma":[0.00007686255,0.0002683964,0.002653982,0.00007720181,0.00002655726,0.0002878512,0.0001370947,0.9141123,0.003992653,0.07594573,0.002384249,0.00003702576],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1855628,0.001087174,0.8032085,0.003343588,0.000151099,0.0003989473,0.0008840843,0.0008327632,0.004531092],"genre_scores_gemma":[0.849373,0.0006717488,0.1451029,0.001088525,0.0001378765,0.0005470241,0.001543273,0.0002213155,0.001314339],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02279486,"threshold_uncertainty_score":0.1205522,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08792836779937473,"score_gpt":0.337564671848105,"score_spread":0.2496363040487302,"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."}}