{"id":"W4283799331","doi":"10.1609/aaai.v36i10.21382","title":"Generation-Focused Table-Based Intermediate Pre-training for Free-Form Question Answering","year":2022,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Topic Modeling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Question answering; Table (database); Leverage (statistics); Schema (genetic algorithms); Artificial intelligence; Natural language processing; Ask price; Language model; Task (project management); Information retrieval; Data mining","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":[0.001327601,0.001804301,0.0008346015,0.0008690518,0.0004450843,0.001095274,0.002476828,0.002065197,0.01243608],"category_scores_gemma":[0.005583252,0.0005337048,0.001549867,0.0007007372,0.000718765,0.002859934,0.00196905,0.003175181,0.004858392],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009264515,"about_ca_system_score_gemma":0.001458344,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003816386,"about_ca_topic_score_gemma":0.007199038,"domain_scores_codex":[0.9990609,0.000310507,0.00005233453,0.0003639974,0.0001094319,0.0001028163],"domain_scores_gemma":[0.9974597,0.001662239,0.00007644696,0.0003937784,0.0002984742,0.0001093439],"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":[0.0007307037,0.0006617621,0.004448628,0.0009129642,0.0001571286,0.0005961311,0.0009928201,0.1405861,0.03023801,0.01226446,0.04961219,0.7587991],"study_design_scores_gemma":[0.0001104588,0.0004400065,0.001046564,0.00008741228,0.00007867702,0.0002850035,0.0002365725,0.9389282,0.01982185,0.02276291,0.01616082,0.00004147423],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0420307,0.00152146,0.9213971,0.0006253482,0.0002827568,0.0005503043,0.002473557,0.02531419,0.005804642],"genre_scores_gemma":[0.4617139,0.000685788,0.501414,0.001531022,0.0001635562,0.001233695,0.02263851,0.001117221,0.00950226],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01243608,"threshold_uncertainty_score":0.04160285,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1391244311080463,"score_gpt":0.3130940979136808,"score_spread":0.1739696668056346,"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."}}