{"id":"W6911194615","doi":"10.5281/zenodo.10104139","title":"Fine-Tuned Large Language Models for Symptom Recognition from Spanish Clinical Text","year":2023,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"","keywords":"Categorization; Task (project management); Set (abstract data type); Generative grammar; Unified Medical Language System; Natural language; Health care; Data curation; Health professionals","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":"codex-gemma-dda1882f352a","candidate_categories":["sts","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.00182061,0.0001564175,0.0002102055,0.0002280871,0.001508865,0.0007801891,0.00168683,0.0001213636,0.001983872],"category_scores_gemma":[0.001823664,0.000169596,0.0001041659,0.0008425234,0.00005568279,0.0004849128,0.001638685,0.0004105378,0.01128543],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007953892,"about_ca_system_score_gemma":0.00001030243,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006208095,"about_ca_topic_score_gemma":0.000002923819,"domain_scores_codex":[0.9972148,0.0006775222,0.0004329874,0.000724217,0.0004054268,0.0005450622],"domain_scores_gemma":[0.9980411,0.0002685016,0.0001602608,0.0007753034,0.0005227366,0.0002320871],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008420695,0.0001891597,0.00005654844,0.0001246894,0.00006389491,0.00004132569,0.004623156,0.0004332497,0.0004727473,0.01415856,0.2460287,0.7337238],"study_design_scores_gemma":[0.001441791,0.0004248725,0.002804761,0.00007824494,0.00001469659,0.00002182431,0.0003324869,0.2897943,0.0001507764,0.007236841,0.6973317,0.0003676531],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1392846,0.0001614513,0.8041765,0.008818096,0.001112341,0.002083346,0.002221439,0.009678489,0.03246371],"genre_scores_gemma":[0.9654229,0.0001144987,0.01622718,0.00102131,0.001019916,7.78325e-7,0.01191112,0.001941225,0.002341071],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8261383,"threshold_uncertainty_score":0.999791,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08905178192975607,"score_gpt":0.3355385511252263,"score_spread":0.2464867691954702,"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."}}