{"id":"W4366462753","doi":"10.1177/08465371231171125","title":"Accuracy of Information and References Using ChatGPT-3 for Retrieval of Clinical Radiological Information","year":2023,"lang":"en","type":"article","venue":"Canadian Association of Radiologists Journal","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":126,"is_retracted":false,"has_abstract":true,"ca_institutions":"Hospital for Sick Children; University of Toronto","funders":"","keywords":"Medicine; Subspecialty; Radiological weapon; Correctness; Information retrieval; Checklist; Medical physics; Radiology; Computer science; Family medicine; Psychology","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":[0.003756889,0.00006683291,0.0003838932,0.0003245299,0.00009556476,0.00001492231,0.00007202022,0.0002803064,0.00001859391],"category_scores_gemma":[0.01903455,0.00005706516,0.000103209,0.0003169725,0.0001004589,0.0004759161,0.000007128911,0.0002487832,0.000002620487],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003086798,"about_ca_system_score_gemma":0.001610222,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002812643,"about_ca_topic_score_gemma":0.0005525469,"domain_scores_codex":[0.9979925,0.0001339021,0.001417173,0.00005730842,0.0001976578,0.0002015287],"domain_scores_gemma":[0.9958394,0.0009969758,0.001808128,0.00007884025,0.001066928,0.0002097683],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0004892715,0.00002293701,0.7703073,0.0002819267,0.0001311939,0.000001011737,0.003104856,0.0002405786,0.0002081231,0.001130795,0.005527431,0.2185546],"study_design_scores_gemma":[0.0008387927,0.001751959,0.9549661,0.0003505609,0.0001551607,0.000112914,0.00623948,0.01282894,0.001732696,0.004134889,0.01669221,0.0001962894],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9956375,0.00009929162,0.001015926,0.002146587,0.0006177816,0.0002891123,0.00007143115,0.000008124297,0.0001142667],"genre_scores_gemma":[0.9966814,0.001422985,0.00138921,0.0001800626,0.0002063576,0.00000195736,0.0001026089,0.000002529191,0.00001288235],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2183583,"threshold_uncertainty_score":0.9892285,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2617026522178948,"score_gpt":0.4714570099247067,"score_spread":0.2097543577068119,"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."}}