{"id":"W4401042733","doi":"10.18653/v1/2024.semeval-1.26","title":"iML at SemEval-2024 Task 2: Safe Biomedical Natural Language Interference for Clinical Trials with LLM Based Ensemble Inferencing","year":2024,"lang":"en","type":"article","venue":"","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"SemEval; Computer science; Task (project management); Natural language processing; Artificial intelligence; Natural language; Interference (communication); Natural (archaeology); Engineering; Biology","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":[0.04043778,0.002651272,0.00198806,0.002384767,0.001766598,0.004007995,0.005683573,0.005588836,0.01674519],"category_scores_gemma":[0.1102791,0.001050039,0.003410593,0.001396539,0.002048287,0.005111119,0.008371103,0.006703523,0.005776833],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003586704,"about_ca_system_score_gemma":0.008456159,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004378188,"about_ca_topic_score_gemma":0.007771085,"domain_scores_codex":[0.9665063,0.02352579,0.00170857,0.004147958,0.003327226,0.0007841443],"domain_scores_gemma":[0.9109158,0.07009332,0.002028782,0.009924797,0.005569712,0.001467603],"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":[0.003801213,0.00173726,0.006985625,0.00638386,0.001649857,0.0025531,0.002604027,0.1174662,0.02372452,0.06476979,0.3040847,0.4642397],"study_design_scores_gemma":[0.001272305,0.0009633601,0.002416929,0.0006445121,0.0006432029,0.001329907,0.0005713479,0.6480435,0.0523715,0.1601529,0.1313336,0.0002568568],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04214981,0.003414886,0.8371621,0.0132691,0.001616771,0.002768194,0.03987087,0.04546418,0.01428398],"genre_scores_gemma":[0.2643345,0.0006770795,0.6550066,0.004400594,0.0006422655,0.002808032,0.06182501,0.003302654,0.00700318],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04043778,"threshold_uncertainty_score":0.213858,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08176553105415232,"score_gpt":0.4291381895587904,"score_spread":0.3473726585046381,"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."}}