{"id":"W4385570307","doi":"10.18653/v1/2023.acl-long.488","title":"DuNST: Dual Noisy Self Training for Semi-Supervised Controllable Text Generation","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Controllability; Text generation; Fluency; Generalization; Artificial intelligence; Language model; Construct (python library); Process (computing); Exploit; Dual (grammatical number); Space (punctuation); Boundary (topology); Natural language processing; Machine learning; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.001852855,0.001160001,0.0009445426,0.0006455139,0.0005342232,0.0008226693,0.002420263,0.001494459,0.002610699],"category_scores_gemma":[0.005469186,0.0006005939,0.0008587754,0.0005054257,0.001180198,0.002224658,0.002199041,0.002182887,0.00126592],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007261287,"about_ca_system_score_gemma":0.0008983317,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002013832,"about_ca_topic_score_gemma":0.004498079,"domain_scores_codex":[0.9989096,0.0004048009,0.0000537387,0.0003849178,0.0001633415,0.00008363075],"domain_scores_gemma":[0.9966691,0.002060148,0.0001893063,0.0006111845,0.0003197231,0.0001505627],"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":[0.0005025725,0.0004659813,0.002585168,0.000335624,0.00012392,0.0002558421,0.0004736619,0.4584861,0.02401453,0.01268468,0.01114758,0.4889245],"study_design_scores_gemma":[0.00001909622,0.00005176767,0.0000910948,0.000006678767,0.000005453245,0.00002411922,0.00001314411,0.991852,0.003113497,0.004003194,0.0008125412,0.000007447718],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03019946,0.0004186537,0.9601125,0.0002517605,0.00009714899,0.0001198698,0.0002198832,0.007058704,0.001522119],"genre_scores_gemma":[0.6122876,0.0002297286,0.3752748,0.0007188923,0.0001754037,0.0005634192,0.002119535,0.001287784,0.007342753],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002610699,"threshold_uncertainty_score":0.009798944,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07735932174670801,"score_gpt":0.2725283510649023,"score_spread":0.1951690293181942,"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."}}