{"id":"W4403603005","doi":"10.2196/60164","title":"Health Care Language Models and Their Fine-Tuning for Information Extraction: Scoping Review","year":2024,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Topic Modeling","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Terminology; Computer science; Health care; Data extraction; Scopus; Unified Medical Language System; Knowledge management; MEDLINE; Data science; Artificial intelligence; Linguistics","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":[],"consensus_categories":[],"category_scores_codex":[0.0005869424,0.0001117716,0.000211725,0.0000937674,0.0001068603,0.0002200168,0.000281276,0.00007378631,0.000007522706],"category_scores_gemma":[0.00007319651,0.00008424347,0.00004791495,0.0001787978,0.00001991503,0.00261042,0.0001423762,0.0002165549,0.00001008855],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005798697,"about_ca_system_score_gemma":0.0004195355,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006430901,"about_ca_topic_score_gemma":0.000008046191,"domain_scores_codex":[0.9987238,0.00001900872,0.0006304967,0.00008394098,0.0003300989,0.0002126412],"domain_scores_gemma":[0.9992627,0.0001472131,0.0001128138,0.0002480783,0.00006130445,0.000167926],"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":[6.927361e-7,0.000002771646,3.754611e-7,0.02548959,0.00000648727,9.036165e-7,0.04679316,0.00008872675,3.421858e-7,0.01494332,0.001712457,0.9109612],"study_design_scores_gemma":[0.0001264356,0.00003225413,5.091633e-7,0.02454061,0.000002426975,0.00004774005,0.001708448,0.9678379,0.000007396788,0.0003195286,0.005276313,0.0001003874],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0002064054,0.02503489,0.969083,0.00335291,0.0002761154,0.0007510217,0.000005569069,0.0002434907,0.001046648],"genre_scores_gemma":[0.2811228,0.02280827,0.6333742,0.05971083,0.001099524,0.001312317,0.0003861969,0.00005759421,0.0001282629],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9677492,"threshold_uncertainty_score":0.3435349,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03758902317528386,"score_gpt":0.3692044030291242,"score_spread":0.3316153798538404,"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."}}