{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001813085,0.0008012204,0.0007248234,0.0004954817,0.0007660373,0.003007741,0.001751449,0.001577203,0.005015324],"category_scores_gemma":[0.01067652,0.0006422367,0.001573874,0.000477709,0.0008687969,0.007828578,0.00153026,0.002147106,0.0007946652],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001004243,"about_ca_system_score_gemma":0.001479653,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01266435,"about_ca_topic_score_gemma":0.02314606,"domain_scores_codex":[0.9993929,0.0003145195,0.00003520787,0.0001186903,0.00009105034,0.00004760709],"domain_scores_gemma":[0.9955949,0.003409914,0.000200512,0.0002723746,0.0003807053,0.0001415806],"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":[0.0005615157,0.0002430667,0.004634447,0.0003991731,0.000231693,0.0006040989,0.001899342,0.5639757,0.002550392,0.3253572,0.0147123,0.08483113],"study_design_scores_gemma":[0.00003487249,0.00002415153,0.0001199606,0.00003036911,0.00004321412,0.00005020259,0.0001158549,0.8421968,0.0006164995,0.1534117,0.003342964,0.00001349501],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06624085,0.001706175,0.9056507,0.006423213,0.0003763445,0.0001720981,0.001666864,0.001235196,0.0165285],"genre_scores_gemma":[0.7345327,0.00153639,0.252381,0.0008991758,0.000200356,0.0002489907,0.002828469,0.0002737608,0.007099261],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01266435,"threshold_uncertainty_score":0.02518129,"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."}}