{"id":"W4385569780","doi":"10.18653/v1/2023.acl-long.307","title":"Evaluating Open-Domain Question Answering in the Era of Large Language Models","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":99,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Matching (statistics); Benchmark (surveying); Question answering; Open domain; Natural language processing; Domain (mathematical analysis); Artificial intelligence; Language model; Information retrieval; Machine learning; Statistics","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":[0.002882064,0.00005023001,0.00008062767,0.00005110252,0.0000446241,0.00008251379,0.0009784786,0.00002174556,0.000007457849],"category_scores_gemma":[0.00004602956,0.00003628817,0.00001731978,0.000424699,0.0000052007,0.0004466597,0.0004395223,0.0001043142,0.000009433638],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001543492,"about_ca_system_score_gemma":0.00003062514,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004624392,"about_ca_topic_score_gemma":0.000186342,"domain_scores_codex":[0.9990897,0.0001508445,0.0001766313,0.0001825335,0.0002330789,0.0001672262],"domain_scores_gemma":[0.9994295,0.00009318264,0.00003762866,0.0004059429,0.00002029523,0.00001340146],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000002745569,0.00004277446,0.0003631002,0.00002315797,0.000004199134,0.00002090608,0.03553703,0.1232293,0.006500541,0.7734329,0.00008736906,0.060756],"study_design_scores_gemma":[0.000197643,0.00001478358,0.0005619857,0.00003284966,5.874398e-7,0.000001611961,0.001033854,0.973222,0.0003653305,0.0245143,0.000009274321,0.00004584352],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4559368,0.00002100047,0.5403343,0.0006596326,0.00004401007,0.0001299034,4.122337e-7,0.00005132568,0.002822661],"genre_scores_gemma":[0.8961189,0.00000235548,0.1035378,0.0002140412,0.00001564512,0.00001557082,0.000001146092,0.000003168171,0.0000913911],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8499926,"threshold_uncertainty_score":0.1818273,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07837739210868339,"score_gpt":0.3860292843682005,"score_spread":0.3076518922595171,"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."}}