{"id":"W3161757350","doi":"10.18653/v1/2021.acl-long.86","title":"MATE-KD: Masked Adversarial TExt, a Companion to Knowledge Distillation","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University; University of British Columbia","funders":"","keywords":"Computer science; Distillation; Benchmark (surveying); Divergence (linguistics); Adversarial system; Generator (circuit theory); Artificial intelligence; Key (lock); Set (abstract data type); Test set; Machine learning; Natural language processing; Ranking (information retrieval); Power (physics); Programming language; Linguistics","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0002806511,0.0002669659,0.0003596974,0.0001613172,0.00008233688,0.0006331019,0.001137708,0.0002286083,0.0001347091],"category_scores_gemma":[0.00005824231,0.0002666052,0.000134485,0.0002339533,0.00001219385,0.0002436355,0.003767569,0.000326254,0.0001981973],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002169452,"about_ca_system_score_gemma":0.0002713061,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000248714,"about_ca_topic_score_gemma":0.0002147081,"domain_scores_codex":[0.9978691,0.0001162194,0.0004152449,0.0009932063,0.0003182924,0.0002878748],"domain_scores_gemma":[0.9981729,0.00006114307,0.0001135114,0.001293328,0.0001956293,0.0001634596],"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":[0.00005867541,0.0006157616,0.001987276,0.001004436,0.000332087,0.0001137804,0.02760396,0.2577363,0.003222671,0.3704535,0.01345258,0.3234189],"study_design_scores_gemma":[0.0002904497,0.00001747445,0.002996492,0.0001773233,0.00001631863,0.000006437179,0.00009179457,0.9851568,0.0004234148,0.002414674,0.007912864,0.0004960048],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02986609,0.00008023655,0.949007,0.001025773,0.004137286,0.0003996019,0.000004252939,0.0004083683,0.01507141],"genre_scores_gemma":[0.7945306,0.000004737704,0.2035471,0.0001230349,0.0004007896,0.00002699461,0.00003426927,0.00001488762,0.001317542],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7646645,"threshold_uncertainty_score":0.9999786,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03359519427592566,"score_gpt":0.2843402056826809,"score_spread":0.2507450114067553,"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."}}