{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006683445,0.0001973709,0.0002599231,0.0001607108,0.00004742268,0.0001384006,0.0006308632,0.0001232825,0.000003397551],"category_scores_gemma":[0.0001571745,0.0002266868,0.00008922064,0.0004786926,0.0001181763,0.0001649794,0.0009912532,0.0002701549,0.00002539129],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001170292,"about_ca_system_score_gemma":0.000238707,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000501936,"about_ca_topic_score_gemma":0.000008750693,"domain_scores_codex":[0.9982558,0.0001496674,0.0004164262,0.0006596758,0.0003635271,0.0001549398],"domain_scores_gemma":[0.998332,0.000248202,0.0003406176,0.0007776056,0.0002078554,0.00009372209],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00001271614,0.00008522894,0.003074975,0.0008404981,0.0001507832,0.00003065351,0.001529011,0.03718581,0.01658404,0.918399,0.006413906,0.01569342],"study_design_scores_gemma":[0.0002294672,0.00002034784,0.09144939,0.0002889108,0.00003790331,0.000002057713,0.00001640561,0.2034239,0.09236921,0.6097415,0.001883091,0.0005377922],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1316412,0.0001083421,0.8628034,0.003033857,0.0009991522,0.0003023109,0.00001200714,0.0002674717,0.0008322402],"genre_scores_gemma":[0.8110165,0.0002142172,0.186704,0.0001157682,0.0002992263,0.0001089461,0.0009083279,0.00004582429,0.00058719],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6793753,"threshold_uncertainty_score":0.9244018,"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."}}