{"id":"W4408244728","doi":"10.2196/65108","title":"ChatGPT’s Performance on Portuguese Medical Examination Questions: Comparative Analysis of ChatGPT-3.5 Turbo and ChatGPT-4o Mini","year":2025,"lang":"en","type":"article","venue":"JMIR Medical Education","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"McNemar's test; Categorical variable; Test (biology); Computer science; Turbo; Portuguese; Consistency (knowledge bases); Medical education; Medical physics; Statistics; Machine learning; Artificial intelligence; Medicine; Mathematics; Engineering","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001227996,0.0002039583,0.0005690965,0.00105214,0.0001363998,0.00001559986,0.0001578775,0.0004010961,0.001027713],"category_scores_gemma":[0.002193344,0.0001783032,0.0001047038,0.001704385,0.0003474355,0.0001619888,0.00003680862,0.0004749885,0.00002992614],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002153726,"about_ca_system_score_gemma":0.002777162,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004691327,"about_ca_topic_score_gemma":0.0001802877,"domain_scores_codex":[0.997129,0.0001876141,0.0008723888,0.0004456507,0.001096723,0.0002686147],"domain_scores_gemma":[0.9978156,0.0004528866,0.0002556452,0.0003996844,0.00057487,0.0005013236],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0002113629,0.002377786,0.05247279,0.0005412435,0.000349108,0.000003159682,0.01259692,0.000006266044,0.00005770567,0.005551651,0.004250173,0.9215818],"study_design_scores_gemma":[0.0002483859,0.0006252066,0.9581839,0.002108237,0.0006431484,0.00001788038,0.006610055,0.02242023,0.002334343,0.0003726854,0.006200631,0.0002352609],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9739252,0.001194381,0.000156415,0.01855587,0.001053344,0.0007267242,0.000003995999,0.00005219524,0.004331916],"genre_scores_gemma":[0.9924446,0.001598979,0.0001181617,0.003419727,0.0003882181,0.0003360335,0.0003188397,0.000009147777,0.00136633],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9213465,"threshold_uncertainty_score":0.9998855,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06824479477615698,"score_gpt":0.4625320639048515,"score_spread":0.3942872691286946,"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."}}