{"id":"W4400046055","doi":"10.1186/s12859-024-05847-x","title":"Can large language models understand molecules?","year":2024,"lang":"en","type":"article","venue":"BMC Bioinformatics","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":40,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Cheminformatics; Computer science; Embedding; Representation (politics); Artificial intelligence; Natural language processing; Field (mathematics); Property (philosophy); Machine learning; Bioinformatics; Biology; Mathematics","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":[0.00191658,0.00112107,0.0008815515,0.0007499826,0.0003491071,0.00216124,0.001449067,0.001267809,0.008480775],"category_scores_gemma":[0.01430381,0.0005312428,0.0014799,0.0007631676,0.0009979407,0.007803898,0.001318218,0.003679162,0.004531405],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008535755,"about_ca_system_score_gemma":0.0009863883,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002723437,"about_ca_topic_score_gemma":0.003558038,"domain_scores_codex":[0.9992606,0.0004071232,0.00003105334,0.0001518509,0.00009694895,0.00005229684],"domain_scores_gemma":[0.9951305,0.003794782,0.0002086278,0.0004736313,0.0002826451,0.0001096706],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006142117,0.00024832,0.005258727,0.001444626,0.0004806776,0.0004184983,0.001028788,0.2760257,0.009174532,0.1710577,0.04676422,0.487484],"study_design_scores_gemma":[0.0000367009,0.00006667856,0.0004041867,0.0001392844,0.00004444711,0.00009469137,0.0001476747,0.7185339,0.002370025,0.2652406,0.0128879,0.00003374271],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02896388,0.004177298,0.9335309,0.01728551,0.0005248895,0.0001027624,0.003121343,0.005372178,0.006921201],"genre_scores_gemma":[0.6044142,0.005976825,0.3635186,0.005198875,0.0006601079,0.0005086666,0.007333449,0.001575846,0.01081341],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008480775,"threshold_uncertainty_score":0.02837104,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01972382980271739,"score_gpt":0.2723903958376214,"score_spread":0.252666566034904,"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."}}