{"id":"W4226003933","doi":"10.18653/v1/2022.acl-long.96","title":"Better Language Model with Hypernym Class Prediction","year":2022,"lang":"en","type":"article","venue":"Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","topic":"Topic Modeling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Microsoft Research","keywords":"Perplexity; Computer science; Artificial intelligence; Transformer; Class (philosophy); Language model; WordNet; Security token; Natural language processing; Machine learning; Context (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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001042885,0.0001532736,0.0002188124,0.00007831767,0.0005957738,0.00006097248,0.001183902,0.00005885119,0.000001744232],"category_scores_gemma":[0.002331982,0.0001179968,0.0001741206,0.0003757953,0.00004480813,0.00009923866,0.000600403,0.000274919,5.022285e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003095941,"about_ca_system_score_gemma":0.0001417719,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002070389,"about_ca_topic_score_gemma":0.000002718768,"domain_scores_codex":[0.997762,0.00002811752,0.0004701504,0.0003204796,0.001161831,0.0002573774],"domain_scores_gemma":[0.9969407,0.000268075,0.001109556,0.0001804281,0.001461193,0.00003999556],"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":[0.00003899432,0.00008532618,0.07806792,0.000100234,0.0001063817,1.323507e-7,0.003242698,0.885057,0.0005668527,0.0308653,0.001559821,0.0003093047],"study_design_scores_gemma":[0.0005711712,0.00009561218,0.00810232,0.00007198232,0.00008200118,0.000003310926,0.0004829607,0.9807698,0.0005571424,0.007455099,0.001643527,0.0001650799],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.913618,0.0001229437,0.05333364,0.007579608,0.003414914,0.002071173,0.001101411,0.0003675982,0.01839074],"genre_scores_gemma":[0.9644597,3.754564e-7,0.03366891,0.0003677445,0.0002336743,0.00004072899,0.000009446203,0.00002120883,0.00119824],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09571276,"threshold_uncertainty_score":0.4811772,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006999110390969953,"score_gpt":0.2098802128432188,"score_spread":0.2028811024522488,"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."}}