{"id":"W4408135179","doi":"10.3389/fdgth.2025.1495040","title":"A simplified retriever to improve accuracy of phenotype normalizations by large language models","year":2025,"lang":"en","type":"article","venue":"Frontiers in Digital Health","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Normalization (sociology); Computer science; Natural language processing; Artificial intelligence; Labrador Retriever; Term (time); Language model; Machine learning; Data mining","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001245594,0.0000908811,0.0001835881,0.0000752759,0.00003479373,0.00001580983,0.0001732985,0.0001131634,0.000001774222],"category_scores_gemma":[0.0003803926,0.00008620501,0.0000373017,0.0002682437,0.00004793884,0.000008397342,0.0001128987,0.00006958467,0.000001028157],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003201867,"about_ca_system_score_gemma":0.000189195,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005287542,"about_ca_topic_score_gemma":0.00002558839,"domain_scores_codex":[0.9991619,0.00001641466,0.0002503355,0.0002247675,0.00008106232,0.0002655096],"domain_scores_gemma":[0.9995661,0.0000158683,0.00006488131,0.0002323303,0.00003789434,0.00008294427],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003770186,0.0004193237,0.0105865,0.0002907109,0.00007749124,0.00000201041,0.001012944,0.0000951097,0.004220617,0.001020491,0.7185884,0.2633094],"study_design_scores_gemma":[0.009239754,0.003470557,0.009062566,0.0006932485,0.00004263284,0.00000396665,0.01342354,0.009273321,0.05649374,0.01559049,0.8809973,0.001708858],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2701224,0.006936675,0.7136155,0.001397726,0.0006911276,0.0005219319,0.001097422,0.00004048494,0.005576662],"genre_scores_gemma":[0.9930701,0.00009937891,0.003433672,0.001418874,0.00002273522,0.00001260409,0.0002920593,0.000008794643,0.001641819],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7229477,"threshold_uncertainty_score":0.3515339,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.007173059222425137,"score_gpt":0.2910441178876113,"score_spread":0.2838710586651862,"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."}}