{"id":"W4413467890","doi":"10.2196/68558","title":"Generative Models and Sentence Transformers for the Recognition and Normalization of Continuous and Discontinuous Phenotype Mentions: Model Development and Evaluation","year":2025,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Preprint; Transformer; Generative grammar; Computer science; Sentence; Artificial intelligence; Coronavirus disease 2019 (COVID-19); Natural language processing; Medicine; Engineering; Pathology; World Wide Web; Electrical engineering","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00407003,0.001988212,0.001225403,0.001461519,0.0005446694,0.001411078,0.002505622,0.001924706,0.0049754],"category_scores_gemma":[0.01047772,0.0009214411,0.002177398,0.0008197188,0.0006911222,0.001962939,0.001334464,0.003422642,0.002701052],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002214219,"about_ca_system_score_gemma":0.001899875,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02359043,"about_ca_topic_score_gemma":0.02707115,"domain_scores_codex":[0.9988356,0.0004810444,0.00007524162,0.0003538472,0.0001520186,0.000102191],"domain_scores_gemma":[0.9926679,0.006003017,0.0001864254,0.000342261,0.0006883058,0.0001122041],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006078176,0.0003199132,0.003773273,0.0002941227,0.0002846098,0.0003000823,0.0002317538,0.7650891,0.003766733,0.005320648,0.006835473,0.2131765],"study_design_scores_gemma":[0.00001303334,0.00002764028,0.000178778,0.000009033555,0.00001784444,0.00002386721,0.000009047245,0.9973317,0.0006902812,0.001398101,0.0002939046,0.000006679946],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1173194,0.002469433,0.8522896,0.001705127,0.0003539961,0.0008548821,0.003564277,0.01647004,0.004973332],"genre_scores_gemma":[0.6489792,0.001142727,0.3303151,0.0006856831,0.0001599349,0.001433164,0.009910889,0.0007227728,0.006650561],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02359043,"threshold_uncertainty_score":0.04690629,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04340515198957854,"score_gpt":0.3216704328808726,"score_spread":0.278265280891294,"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."}}