{"id":"W4388025354","doi":"10.2196/51523","title":"Evaluating Large Language Models for the National Premedical Exam in India: Comparative Analysis of GPT-3.5, GPT-4, and Bard","year":2023,"lang":"en","type":"article","venue":"JMIR Medical Education","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":57,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Mainstream; Test (biology); Artificial intelligence; Psychology; Mathematics education; Computer science; Political science; Biology; Ecology","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007587931,0.001733634,0.0008841821,0.002256924,0.0006513039,0.001777061,0.002146096,0.001553525,0.001592493],"category_scores_gemma":[0.01875281,0.0004655258,0.001789293,0.001470249,0.0008011395,0.001635669,0.002151374,0.002394297,0.001150085],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003427147,"about_ca_system_score_gemma":0.002450637,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.05480896,"about_ca_topic_score_gemma":0.05204337,"domain_scores_codex":[0.9959456,0.002099185,0.0004015306,0.0006175014,0.0006418608,0.000294332],"domain_scores_gemma":[0.9823952,0.0129408,0.0006099361,0.001358583,0.001937967,0.0007575266],"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":[0.006623481,0.003907915,0.1649617,0.002125907,0.00196772,0.0009112813,0.00233908,0.4811098,0.00660854,0.002456907,0.0240679,0.3029197],"study_design_scores_gemma":[0.0003633684,0.004588529,0.115815,0.0001751927,0.0007261198,0.0004347469,0.002375079,0.8567562,0.009464465,0.00199328,0.007064454,0.0002435509],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9769152,0.001278513,0.006774076,0.001162012,0.0002060235,0.0002493622,0.004452738,0.003177542,0.005784424],"genre_scores_gemma":[0.9761211,0.0004139758,0.009936603,0.0002938121,0.0000559032,0.0001809449,0.01080833,0.0001382081,0.002051245],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.05480896,"threshold_uncertainty_score":0.1089798,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2911824033239588,"score_gpt":0.5812372745878501,"score_spread":0.2900548712638913,"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."}}