{"id":"W4389519881","doi":"10.18653/v1/2023.findings-emnlp.542","title":"Knowledge-Selective Pretraining for Attribute Value Extraction","year":2023,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Zhàng; Value (mathematics); Natural language processing; Computer science; Artificial intelligence; Philosophy; Linguistics; Psychology; Cognitive science; History; China; Machine learning; Archaeology","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":[0.0007831487,0.001171371,0.0009718214,0.001662886,0.0008460472,0.001082433,0.001964228,0.001159554,0.01077558],"category_scores_gemma":[0.002705335,0.0006803442,0.001105786,0.002047277,0.0005342657,0.002785687,0.001560196,0.002411272,0.006531665],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007706478,"about_ca_system_score_gemma":0.001977882,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01037763,"about_ca_topic_score_gemma":0.0222139,"domain_scores_codex":[0.9993363,0.00008627255,0.00005822792,0.0002857332,0.0001119398,0.0001214024],"domain_scores_gemma":[0.9987967,0.000537569,0.00005832077,0.0002838352,0.0002704286,0.00005307989],"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":[0.0002744844,0.0002620185,0.001861531,0.0001685798,0.00008874339,0.0001417366,0.0001329262,0.01214388,0.02103684,0.003380704,0.02789402,0.9326146],"study_design_scores_gemma":[0.00007532625,0.000194493,0.004140501,0.0001098133,0.0001895698,0.0002575587,0.0002903251,0.8808985,0.06828251,0.01995844,0.02554541,0.00005747551],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06470673,0.001920196,0.8936775,0.0007308092,0.0004504984,0.0003103055,0.002607001,0.02311012,0.01248677],"genre_scores_gemma":[0.4073205,0.001308802,0.5580822,0.0007363383,0.0002210094,0.0004840655,0.0154545,0.0007618267,0.01563068],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01077558,"threshold_uncertainty_score":0.03604788,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04115544669185017,"score_gpt":0.3580614645553213,"score_spread":0.3169060178634712,"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."}}