{"id":"W4385570025","doi":"10.18653/v1/2023.acl-long.385","title":"Few-shot In-context Learning on Knowledge Base Question Answering","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":57,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Knowledge base; Question answering; Executable; Schema (genetic algorithms); Artificial intelligence; Context (archaeology); Natural language; Natural language processing; Entity linking; Natural language understanding; Matching (statistics); Baseline (sea); Information retrieval; Programming language","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":[0.002658963,0.001873252,0.001669752,0.001834539,0.000984301,0.001542357,0.003835572,0.002998902,0.00494409],"category_scores_gemma":[0.009699998,0.0008786952,0.001628544,0.001375878,0.001137983,0.005027267,0.0027261,0.003730817,0.002045196],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001460379,"about_ca_system_score_gemma":0.001176246,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01295598,"about_ca_topic_score_gemma":0.0182312,"domain_scores_codex":[0.9979544,0.0007642198,0.0001050233,0.0008019946,0.0002222075,0.0001521234],"domain_scores_gemma":[0.9953797,0.003229959,0.0001263293,0.000632778,0.0004251843,0.0002060864],"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":[0.0006556789,0.0007666497,0.003682046,0.001176715,0.0003834473,0.0005269257,0.001048686,0.2394385,0.01240421,0.01144164,0.02936997,0.6991055],"study_design_scores_gemma":[0.00004806577,0.0001184113,0.0006364501,0.0000516192,0.00006135921,0.0001371561,0.0001417436,0.9683071,0.003608251,0.02183443,0.005027392,0.00002804144],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05328493,0.005056654,0.9172727,0.00148818,0.0002969436,0.000389106,0.001566659,0.01684937,0.00379543],"genre_scores_gemma":[0.5942798,0.001137568,0.3874054,0.001883056,0.0004024617,0.0004733609,0.008382041,0.0007100949,0.0053262],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01295598,"threshold_uncertainty_score":0.02576119,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05001398167547733,"score_gpt":0.3010416719042726,"score_spread":0.2510276902287952,"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."}}