{"id":"W4409712065","doi":"10.1002/jac5.70038","title":"Accuracy and reproducibility of <scp>ChatGPT</scp> responses to real‐world drug information questions","year":2025,"lang":"en","type":"article","venue":"JACCP JOURNAL OF THE AMERICAN COLLEGE OF CLINICAL PHARMACY","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute for Clinical Evaluative Sciences; University of Toronto","funders":"","keywords":"Reproducibility; Drug; Computer science; Pharmacology; Chemistry; Chromatography; Medicine","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.006389471,0.0001093014,0.0007217769,0.0003507758,0.0000975173,0.000009616761,0.0002812297,0.00002672812,0.00001419008],"category_scores_gemma":[0.0565436,0.00007702297,0.0002564202,0.001250608,0.00043505,0.0002976237,0.0001114142,0.0004892848,0.000003296734],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007240457,"about_ca_system_score_gemma":0.001652365,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001351741,"about_ca_topic_score_gemma":0.00007710755,"domain_scores_codex":[0.9960803,0.0006240649,0.002576198,0.000217263,0.0003459189,0.0001562067],"domain_scores_gemma":[0.986419,0.007780582,0.002848272,0.0009847756,0.001694229,0.000273097],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.002848321,0.0005698421,0.7372982,0.0002326246,0.0001616083,0.000003584862,0.001244753,0.00004546571,0.00107772,0.0003972786,0.06825458,0.187866],"study_design_scores_gemma":[0.0005994019,0.0007853052,0.8096926,0.000866174,0.0004102907,0.00003838403,0.004507144,0.0001956068,0.04184771,0.005012152,0.1359752,0.00007002264],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9556549,0.0001934837,0.00009852819,0.04153806,0.001156845,0.0004915237,0.0000324402,0.000008221683,0.0008260029],"genre_scores_gemma":[0.9924083,0.00151122,0.001577027,0.002862445,0.0002961915,0.000005165495,8.687618e-7,0.00000613264,0.001332598],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.187796,"threshold_uncertainty_score":0.9514036,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1454278925233091,"score_gpt":0.5277380696326105,"score_spread":0.3823101771093014,"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."}}