{"id":"W4411118750","doi":"10.1016/j.nuclcard.2025.102269","title":"Priming with specific context improves large language model performance on nuclear cardiology board preparation test","year":2025,"lang":"en","type":"article","venue":"Journal of Nuclear Cardiology","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"National Heart, Lung, and Blood Institute; National Institutes of Health","keywords":"Medicine; Context (archaeology); Test (biology); Internal medicine; Cardiology; Medical physics","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":[0.0004770334,0.0001453985,0.0005844588,0.000223374,0.0001518148,0.00001664206,0.0001203046,0.0001961274,0.0000185094],"category_scores_gemma":[0.0002088923,0.0001107331,0.0001606672,0.0001339424,0.000109475,0.0001203234,0.00002716836,0.0005075842,0.00004555808],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002051793,"about_ca_system_score_gemma":0.0002232682,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001648377,"about_ca_topic_score_gemma":0.000006568286,"domain_scores_codex":[0.9987332,0.00009296974,0.0005157643,0.0002079724,0.0001595766,0.0002905209],"domain_scores_gemma":[0.9988742,0.0001659352,0.0002495755,0.0002768013,0.0003353868,0.00009809983],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.02915788,0.001185467,0.2873143,0.001235804,0.00260487,0.001328343,0.0356366,0.03288097,0.1520611,0.01308687,0.1642006,0.2793072],"study_design_scores_gemma":[0.004375707,0.04038673,0.3534194,0.002690416,0.001703796,0.005398491,0.08594721,0.07801447,0.009465915,0.0007468605,0.4163256,0.001525388],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9910231,0.0003568808,0.0004879998,0.00169969,0.000569388,0.0002350191,0.000003316983,0.00002760923,0.005597061],"genre_scores_gemma":[0.9962435,0.0004689546,0.0005932664,0.001686676,0.0007785624,0.000003051051,0.000003342326,0.00002553612,0.0001970637],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2777818,"threshold_uncertainty_score":0.4515565,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05156742709689905,"score_gpt":0.3576561139659181,"score_spread":0.3060886868690191,"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."}}