{"id":"W3161374759","doi":"10.18653/v1/2021.naacl-main.102","title":"Understanding by Understanding Not: Modeling Negation in Language Models","year":2021,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":40,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Microsoft (Canada); Université de Montréal; Montreal Police Service","funders":"","keywords":"Negation; Computer science; Computational linguistics; Linguistics; Cognitive science; Programming language; Natural language processing; Artificial intelligence; Psychology; Philosophy","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":[0.0002788508,0.0001258012,0.0001390588,0.0001541221,0.0001009464,0.0003162657,0.0003856748,0.00009138735,0.00001172787],"category_scores_gemma":[0.0000468107,0.0001231977,0.00003656455,0.0006383762,0.00001432159,0.001341537,0.0002106112,0.0002079845,0.000002322246],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007576362,"about_ca_system_score_gemma":0.00008018585,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000100464,"about_ca_topic_score_gemma":0.0001287128,"domain_scores_codex":[0.9987761,0.00005145064,0.0002194747,0.0003931056,0.0002841643,0.0002756969],"domain_scores_gemma":[0.999486,0.00007063907,0.00004617715,0.0003102837,0.00003516028,0.00005170196],"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":[0.000002946621,0.00002061535,0.00001016524,0.00001947726,0.000004570124,0.00005773829,0.001924318,0.005451564,0.01587961,0.9754133,0.0002679284,0.0009478275],"study_design_scores_gemma":[0.0001169972,0.000004437249,6.907204e-8,0.00004435173,0.000001238872,0.000008321652,0.001235333,0.6506559,0.01974009,0.3280748,0.000001404778,0.0001171471],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001510264,0.0009465732,0.9910589,0.001191887,0.00007075674,0.0000726968,0.000001246085,0.0005228501,0.004624866],"genre_scores_gemma":[0.7548098,0.00001757171,0.2445281,0.0004470143,0.00001256939,0.000003555387,0.000006132218,0.000009452843,0.0001658621],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7532995,"threshold_uncertainty_score":0.5023856,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1277660868719435,"score_gpt":0.2968482960916106,"score_spread":0.1690822092196671,"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."}}