{"id":"W3183195595","doi":"10.1162/tacl_a_00386","title":"Context-aware Adversarial Training for Name Regularity Bias in Named Entity Recognition","year":2021,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Huawei Technologies (Canada)","funders":"","keywords":"Adversarial system; Focus (optics); Testbed; Named-entity recognition; Training set; Noise (video); Training (meteorology)","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":[0.000693304,0.00009372538,0.0001970142,0.00007607152,0.000228355,0.00005819464,0.0002614815,0.0001111131,0.000004773474],"category_scores_gemma":[0.003625738,0.0001017544,0.0002179209,0.0002659278,0.00001916701,0.00008954848,0.00001602497,0.0001419275,0.00000108941],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002959526,"about_ca_system_score_gemma":0.0004013327,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003958949,"about_ca_topic_score_gemma":0.0001764161,"domain_scores_codex":[0.998693,0.000108775,0.0004549617,0.0002446176,0.0003250575,0.0001735547],"domain_scores_gemma":[0.9964706,0.001399694,0.0003380431,0.0001794613,0.001581309,0.00003085154],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001066852,0.0005314706,0.002938704,0.0002225809,0.0003277532,0.000001426783,0.004344951,0.7570647,0.00007458376,0.1905416,0.0003083444,0.04353723],"study_design_scores_gemma":[0.001680068,0.00003212303,0.001833383,0.00006165829,0.00006014421,0.000001559263,0.0001703969,0.8222667,0.0005645643,0.1707774,0.002406139,0.0001458815],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004614749,0.00001795619,0.9916773,0.0009005243,0.001958403,0.0003669694,0.0002951966,0.00004263799,0.0001262508],"genre_scores_gemma":[0.8654615,0.000001870407,0.133945,0.00009927167,0.0001644267,0.00003726227,0.00009067329,0.000007903566,0.000192093],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8608468,"threshold_uncertainty_score":0.4340609,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0861111343750003,"score_gpt":0.2954547619652698,"score_spread":0.2093436275902695,"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."}}