{"id":"W4399197789","doi":"10.48550/arxiv.2405.18448","title":"Multi-objective Representation for Numbers in Clinical Narratives: A CamemBERT-Bio-Based Alternative to Large-Scale LLMs","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Mathematics, Computing, and Information Processing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Fonds de Recherche du Québec - Santé; Institut de Valorisation des Données; Université de Montréal","keywords":"Representation (politics); Narrative; Mathematics; Forestry; Geography; Art; Literature; Political science","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"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":[0.001005843,0.001139955,0.0004378704,0.001368817,0.0002305447,0.001300228,0.001258345,0.0009971168,0.006785862],"category_scores_gemma":[0.005462193,0.0002413223,0.0009156813,0.0008657603,0.0004225524,0.002690035,0.001520457,0.001497161,0.003138796],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009651345,"about_ca_system_score_gemma":0.0009029812,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003114254,"about_ca_topic_score_gemma":0.006276061,"domain_scores_codex":[0.9993544,0.0002247181,0.00005653146,0.0002122115,0.0001178083,0.000034207],"domain_scores_gemma":[0.9985147,0.000765211,0.0001515593,0.0002709328,0.0002332557,0.00006429334],"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":[0.000746535,0.0003455747,0.005921473,0.0008124206,0.0001818402,0.0004512363,0.0007496284,0.07144203,0.01815258,0.02478552,0.03478713,0.841624],"study_design_scores_gemma":[0.00006295054,0.0002272967,0.002102893,0.0001426303,0.00005396347,0.0002233509,0.0003209315,0.9170781,0.01358148,0.03091102,0.03524012,0.00005520462],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04612789,0.002399998,0.9160665,0.002205344,0.0004820543,0.0003631758,0.00977563,0.01616168,0.006417648],"genre_scores_gemma":[0.3793007,0.00113727,0.5844852,0.001213705,0.0002796488,0.0006351366,0.02090121,0.0009104679,0.01113667],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006785862,"threshold_uncertainty_score":0.02270103,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1069859715520276,"score_gpt":0.2959768351915429,"score_spread":0.1889908636395153,"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."}}