{"id":"W4412791970","doi":"10.1001/jamaophthalmol.2025.2413","title":"Ophthalmological Question Answering and Reasoning Using OpenAI o1 vs Other Large Language Models","year":2025,"lang":"en","type":"article","venue":"JAMA Ophthalmology","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"U.S. National Library of Medicine","keywords":"Metric (unit); Medicine; Macro; Artificial intelligence; Natural language processing; Machine learning; Computer science; Operations management","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":[0.0006065304,0.0001525189,0.0003165807,0.0001724028,0.0001789921,0.00003790393,0.00008529668,0.0003300952,0.0002495871],"category_scores_gemma":[0.000428682,0.0001385773,0.00004843407,0.0002356738,0.00009202299,0.0001873468,0.00008784929,0.0003296982,0.00001585727],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001162208,"about_ca_system_score_gemma":0.000141739,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002224351,"about_ca_topic_score_gemma":0.000007374359,"domain_scores_codex":[0.9986317,0.0001635905,0.0003417617,0.0003760043,0.0001049878,0.0003820181],"domain_scores_gemma":[0.9992752,0.0001629321,0.00007760672,0.000255405,0.0001168217,0.0001120253],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002160589,0.0009631833,0.872445,0.00088752,0.0002029217,0.003475001,0.008336138,0.001006222,0.01938253,0.05463227,0.000572873,0.0359357],"study_design_scores_gemma":[0.002315906,0.003729707,0.3540471,0.00457823,0.0006616513,0.06456158,0.0274416,0.4248398,0.04206912,0.07106282,0.002742046,0.001950503],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9827561,0.001109777,0.002796241,0.001771995,0.0004327076,0.0003437686,0.000004794658,0.00005255792,0.01073201],"genre_scores_gemma":[0.9939957,0.00002410169,0.003784279,0.0008905069,0.0002325866,0.00002686336,0.00001513694,0.00001801432,0.001012773],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.518398,"threshold_uncertainty_score":0.5651019,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1228985104852794,"score_gpt":0.4578100303736842,"score_spread":0.3349115198884048,"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."}}