{"id":"W4391644915","doi":"10.2196/50965","title":"Comparison of the Performance of GPT-3.5 and GPT-4 With That of Medical Students on the Written German Medical Licensing Examination: Observational Study","year":2024,"lang":"en","type":"article","venue":"JMIR Medical Education","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":104,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"German; Enthusiasm; Observational study; Medical education; Medicine; Medical care; Psychology; Family medicine; Internal medicine","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002453804,0.000121119,0.0002948105,0.000107129,0.0001051259,0.00001840341,0.0003754165,0.0001946501,0.0006329322],"category_scores_gemma":[0.001269213,0.00006452961,0.00004416612,0.0004643875,0.0004593954,0.00008339533,0.00008245606,0.0005538937,0.000004419496],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006891723,"about_ca_system_score_gemma":0.002693009,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002315835,"about_ca_topic_score_gemma":0.00009105761,"domain_scores_codex":[0.9943179,0.0002232618,0.0007000853,0.0002262678,0.004388378,0.0001441382],"domain_scores_gemma":[0.9980978,0.0008460798,0.0002132586,0.0003048935,0.0003195351,0.0002184453],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00008703976,0.002474986,0.8623416,0.000597815,0.00006661213,0.000001584425,0.02101723,8.70438e-7,0.00002671086,0.001151796,0.002166576,0.1100671],"study_design_scores_gemma":[0.0001420763,0.0008069503,0.9726672,0.003465933,0.00007644387,0.00003115981,0.01408318,0.006192499,0.001650714,0.00005006156,0.000763359,0.00007046653],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9523399,0.0003011078,0.00001734286,0.04571363,0.0006571569,0.0007924744,0.000001058239,0.00001389488,0.0001633892],"genre_scores_gemma":[0.9981716,0.0001034897,0.00002143034,0.001100457,0.0003835984,0.00008504235,0.00001572079,0.00001261352,0.0001059923],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1103255,"threshold_uncertainty_score":0.6930164,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2296316867141716,"score_gpt":0.5298358915005704,"score_spread":0.3002042047863989,"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."}}