{"id":"W4402670684","doi":"10.18653/v1/2024.findings-acl.151","title":"Accurate and Nuanced Open-QA Evaluation Through Textual Entailment","year":2024,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Textual entailment; Computer science; Natural language processing; Logical consequence; Artificial intelligence; Information retrieval","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":[0.009448487,0.001692873,0.001356539,0.00332161,0.001996337,0.004783839,0.002473661,0.003489592,0.01958059],"category_scores_gemma":[0.05501971,0.0006184696,0.001130294,0.001668914,0.001247976,0.006530402,0.005158877,0.002784559,0.009476873],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001939615,"about_ca_system_score_gemma":0.002973776,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00481062,"about_ca_topic_score_gemma":0.006868142,"domain_scores_codex":[0.9843138,0.008147961,0.001145261,0.00199621,0.003714147,0.0006826414],"domain_scores_gemma":[0.9670236,0.01729575,0.001177025,0.004438859,0.009223205,0.0008416212],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.002653688,0.001089713,0.005941579,0.002787152,0.0003981147,0.0009917759,0.002147296,0.03735646,0.06951284,0.02913031,0.09447676,0.7535143],"study_design_scores_gemma":[0.0004755436,0.0008343671,0.006687524,0.0008138312,0.0004120973,0.001286884,0.002003842,0.7131881,0.1006793,0.09296554,0.08040881,0.0002441189],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1214807,0.004366569,0.797788,0.003077929,0.001386195,0.00105147,0.00766826,0.02862156,0.03455937],"genre_scores_gemma":[0.6137265,0.0006010766,0.3517466,0.0009378454,0.0004604933,0.0003186041,0.01579099,0.002500759,0.01391717],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01958059,"threshold_uncertainty_score":0.06550366,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0861146672899413,"score_gpt":0.3814566730101803,"score_spread":0.2953420057202391,"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."}}