{"id":"W4415299179","doi":"10.3390/make7040121","title":"Small or Large? Zero-Shot or Finetuned? Guiding Language Model Choice for Specialized Applications in Healthcare","year":2025,"lang":"en","type":"article","venue":"Machine Learning and Knowledge Extraction","topic":"Topic Modeling","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Provincial Health Services Authority; University of British Columbia","funders":"","keywords":"Task (project management); Language model; Selection (genetic algorithm); Exploit; Health care; Language understanding; Task analysis","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":[0.003783945,0.001576215,0.0008711349,0.0005919966,0.000362248,0.001664049,0.001761123,0.001477713,0.004088549],"category_scores_gemma":[0.01685872,0.000573346,0.0007387291,0.0003585084,0.0006155013,0.00300407,0.001765345,0.002949166,0.003030244],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008824739,"about_ca_system_score_gemma":0.001969049,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005972823,"about_ca_topic_score_gemma":0.01054606,"domain_scores_codex":[0.9982086,0.0009829272,0.000091699,0.0004553378,0.0001212464,0.0001400426],"domain_scores_gemma":[0.995357,0.003573485,0.0001658778,0.0003785491,0.0003282829,0.0001968102],"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":[0.001628721,0.0007786423,0.01229609,0.001123073,0.0003073908,0.0004102984,0.0009228894,0.1659441,0.03834131,0.003679191,0.02431096,0.7502574],"study_design_scores_gemma":[0.0001917364,0.0005317809,0.002488552,0.0002194222,0.0001657302,0.0002486137,0.0005369284,0.9524067,0.02291664,0.01139117,0.008829094,0.00007356532],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3671596,0.008005193,0.5793918,0.006596797,0.0004926883,0.0006141945,0.001565127,0.02491004,0.01126445],"genre_scores_gemma":[0.8342974,0.0008828363,0.1550638,0.001937551,0.00009620319,0.0003591032,0.002016504,0.001058127,0.004288551],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005972823,"threshold_uncertainty_score":0.0200116,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07084545954893032,"score_gpt":0.3752476515817091,"score_spread":0.3044021920327787,"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."}}