{"id":"W4385573653","doi":"10.18653/v1/2022.emnlp-main.696","title":"Dictionary-Assisted Supervised Contrastive Learning","year":2022,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"York University; Charles Koch Foundation; Craig Newmark Philanthropies; John S. and James L. Knight Foundation; William and Flora Hewlett Foundation; Bill and Melinda Gates Foundation","keywords":"Computer science; Leverage (statistics); Natural language processing; Artificial intelligence; Entropy (arrow of time); Word (group theory); Cross entropy; Security token; Principle of maximum entropy; Linguistics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002993062,0.00141532,0.001362111,0.001439209,0.0007404901,0.001201897,0.002195975,0.001463137,0.002218093],"category_scores_gemma":[0.01090881,0.0004300515,0.00105764,0.001131411,0.001448959,0.002645425,0.002748589,0.002827993,0.001379512],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000748013,"about_ca_system_score_gemma":0.000980039,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001319547,"about_ca_topic_score_gemma":0.003235851,"domain_scores_codex":[0.9979469,0.0007748479,0.0001203391,0.0007274716,0.0003132854,0.0001170858],"domain_scores_gemma":[0.9934927,0.00396139,0.000451108,0.0009986043,0.0009219297,0.0001743188],"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":[0.0008653092,0.0008151134,0.005538705,0.0004388269,0.0003092388,0.0002552262,0.0005999707,0.2213287,0.0247072,0.01417511,0.01167939,0.7192872],"study_design_scores_gemma":[0.00003578552,0.0001398686,0.0004500716,0.00001511329,0.00002099081,0.00005975234,0.00004697934,0.981586,0.005950934,0.01014263,0.001534997,0.0000167182],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04470052,0.0006415599,0.9498124,0.0002776477,0.0001215524,0.0001217809,0.0003181165,0.002291201,0.001715218],"genre_scores_gemma":[0.606303,0.0003054355,0.3827296,0.0006482087,0.0002920497,0.0004441699,0.003114459,0.0005387575,0.00562432],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002993062,"threshold_uncertainty_score":0.01582903,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02082510238209005,"score_gpt":0.2238134163181582,"score_spread":0.2029883139360682,"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."}}