{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01285615,0.002883431,0.001314166,0.003055743,0.001204005,0.002874946,0.002531046,0.002646512,0.003888877],"category_scores_gemma":[0.05178178,0.0005784282,0.001092024,0.001841704,0.001409472,0.003660867,0.004786846,0.003014585,0.003333192],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001877344,"about_ca_system_score_gemma":0.001374885,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005043851,"about_ca_topic_score_gemma":0.007288284,"domain_scores_codex":[0.9777314,0.01477629,0.001103936,0.003178342,0.002483393,0.0007266432],"domain_scores_gemma":[0.9645017,0.02167398,0.001637895,0.005679882,0.004821747,0.001684825],"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":[0.007725429,0.004401443,0.05878584,0.0056298,0.001587345,0.000884786,0.005734972,0.2126558,0.0300513,0.008207325,0.1264062,0.5379299],"study_design_scores_gemma":[0.001029247,0.005047627,0.05493587,0.0007914133,0.0003916781,0.0007555475,0.004108924,0.7847366,0.05995833,0.02215948,0.06562322,0.000462079],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.73103,0.01103184,0.1562697,0.001589743,0.001599192,0.001922347,0.02971164,0.04240571,0.02443974],"genre_scores_gemma":[0.8645595,0.0006093174,0.06819958,0.000445379,0.0002618006,0.001152592,0.05792588,0.001590389,0.005255618],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01285615,"threshold_uncertainty_score":0.06799066,"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."}}