{"id":"W4413289814","doi":"10.26434/chemrxiv-2025-dl4sb","title":"Capturing Electron Correlation with Machine Learning through a Data-Driven CASPT2 Framework","year":2025,"lang":"en","type":"article","venue":"ChemRxiv","topic":"Electron and X-Ray Spectroscopy Techniques","field":"Materials Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Division of Chemistry; National Science Foundation","keywords":"Computer science; Correlation; Electron; Artificial intelligence; Machine learning; Pattern recognition (psychology); Physics; Mathematics; Geometry; Nuclear physics","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":[],"consensus_categories":[],"category_scores_codex":[0.0002339943,0.000191446,0.0002260965,0.00006002347,0.0002233866,0.0001068008,0.0004669157,0.0001236014,0.0001799796],"category_scores_gemma":[0.0001001112,0.0001605037,0.00002722771,0.0003166328,0.00007801193,0.0003530353,0.000149693,0.0005243946,0.00004562082],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001185318,"about_ca_system_score_gemma":0.0000982138,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002475027,"about_ca_topic_score_gemma":0.0001082222,"domain_scores_codex":[0.9986439,0.00005538379,0.0001967033,0.0005061897,0.0001863541,0.0004115079],"domain_scores_gemma":[0.9991302,0.00008620926,0.0001074297,0.0005941382,0.00004473492,0.00003730004],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001708771,0.00009195902,0.00790388,0.00007214687,0.0000393285,0.00001637686,0.0006556933,0.0004772483,0.9580536,0.03084456,0.001360544,0.0003137175],"study_design_scores_gemma":[0.0002435629,0.000171673,0.0003747613,0.0001820799,0.00005123988,0.00001023607,0.00004505143,0.004674874,0.9814904,0.005292275,0.007211353,0.0002524639],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7615274,0.00112241,0.2225311,0.0008264004,0.0001757975,0.0003446095,0.000003205238,0.0009467699,0.01252233],"genre_scores_gemma":[0.9674619,0.00008932059,0.03080397,0.0002608596,0.00007640085,0.0000263299,0.0000640407,0.00002519694,0.001192002],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2059345,"threshold_uncertainty_score":0.6545153,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01531066456444434,"score_gpt":0.2931409582012542,"score_spread":0.2778302936368098,"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."}}