{"id":"W4383368470","doi":"10.31219/osf.io/ry6a4","title":"Training and Calibration of ChatGPT for Reliable and Accurate Assessment","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Calibration; Training (meteorology); Computer science; Quantitative assessment; Artificial intelligence; Reliability engineering; Statistics; Engineering; Mathematics; Geography","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007004644,0.001568779,0.0008535978,0.001306472,0.0006117749,0.001598534,0.002580421,0.002410147,0.01792846],"category_scores_gemma":[0.0586545,0.0007806999,0.0005992019,0.0005902312,0.0006447118,0.002906885,0.004253026,0.002530384,0.01436036],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006570952,"about_ca_system_score_gemma":0.001694317,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002735132,"about_ca_topic_score_gemma":0.001815929,"domain_scores_codex":[0.994186,0.00224172,0.0004809914,0.001083918,0.001598403,0.0004089326],"domain_scores_gemma":[0.9705496,0.01347438,0.0008936028,0.004636558,0.00926702,0.001178735],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002117416,0.0008304961,0.01625908,0.0008004688,0.0001308455,0.0007513679,0.002217176,0.03243573,0.07847801,0.004622898,0.03320822,0.8281484],"study_design_scores_gemma":[0.0003233182,0.000974979,0.02921337,0.0005487218,0.0001429646,0.00213235,0.0009482845,0.7737315,0.1262299,0.01427414,0.05121987,0.0002605324],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05574692,0.0003475605,0.9051675,0.0005016292,0.0006650281,0.0008901244,0.0006268706,0.03032794,0.005726365],"genre_scores_gemma":[0.4042225,0.000223356,0.5806577,0.0005868736,0.0002302021,0.00199945,0.001960685,0.002575913,0.007543318],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01792846,"threshold_uncertainty_score":0.05997664,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.474630810375702,"score_gpt":0.5197450193923994,"score_spread":0.04511420901669738,"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."}}