{"id":"W4296711106","doi":"10.1162/tacl_a_00506","title":"Evaluating Attribution in Dialogue Systems: The BEGIN Benchmark","year":2022,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":44,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Spurious relationship; Benchmark (surveying); Attribution; Artificial intelligence; Natural language processing; Grounded theory; Data science; Machine learning; Qualitative research","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":[],"consensus_categories":[],"category_scores_codex":[0.001894087,0.00007497027,0.0001235328,0.00008083235,0.0006922737,0.00004935094,0.0006068226,0.00003413863,0.000003645364],"category_scores_gemma":[0.0009724029,0.00006319057,0.0001121509,0.0004569704,0.00001601402,0.00004386479,0.00004913532,0.0002188384,8.195505e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006081125,"about_ca_system_score_gemma":0.0002137584,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001125065,"about_ca_topic_score_gemma":0.00001866396,"domain_scores_codex":[0.9982984,0.0003008375,0.0004228133,0.0001732891,0.0006492363,0.0001554076],"domain_scores_gemma":[0.9974352,0.001385207,0.0003938719,0.0002358369,0.0005340975,0.00001582378],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000003683305,0.00003921397,0.0007819016,0.000009559121,0.00001896703,5.361912e-8,0.0003461091,0.8677119,0.000005948008,0.1305601,0.00007326783,0.0004493622],"study_design_scores_gemma":[0.0003558997,0.00004000152,0.003093898,0.000009873927,0.00002460541,0.000001259452,0.00008095613,0.9683258,0.00001170822,0.02505128,0.002936021,0.0000687051],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004118812,0.0000582131,0.9912609,0.001261886,0.002447841,0.0005063805,0.0001513132,0.0000319609,0.0001627247],"genre_scores_gemma":[0.9821219,9.344745e-7,0.01730566,0.00006905398,0.0001326267,0.0001294092,0.00003198841,0.000005957642,0.0002024882],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9780031,"threshold_uncertainty_score":0.5324479,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04673618613088275,"score_gpt":0.3095691801184852,"score_spread":0.2628329939876024,"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."}}