{"id":"W3037585139","doi":"10.18653/v1/2020.repl4nlp-1.10","title":"Exploring the Limits of Simple Learners in Knowledge Distillation for Document Classification with DocBERT","year":2020,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Distillation; Computer science; Task (project management); Baseline (sea); Artificial intelligence; FLOPS; Machine learning; Natural language processing; Parallel computing; Engineering","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001303895,0.00004644306,0.00006300348,0.00002342265,0.00002984322,0.00002868502,0.0002243869,0.000009083259,0.000001482036],"category_scores_gemma":[0.00003369137,0.00002936534,0.00001584322,0.0001957619,0.00001035704,0.0002945818,0.00003829254,0.0000364386,0.000002165715],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002382687,"about_ca_system_score_gemma":0.00002666425,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002599424,"about_ca_topic_score_gemma":0.0000772955,"domain_scores_codex":[0.999496,0.00002148728,0.0001417664,0.0001692736,0.00008497768,0.00008651899],"domain_scores_gemma":[0.9996247,0.00009303275,0.00004379206,0.0001778483,0.00003505455,0.0000255143],"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":[0.00003559366,0.00003189136,0.00573132,0.00006668545,0.00001196596,3.101929e-7,0.01701564,0.02376024,0.001764198,0.6345009,0.0001062271,0.316975],"study_design_scores_gemma":[0.0002768184,0.00007797569,0.008819956,0.00001207383,0.000002620514,2.479906e-7,0.000568777,0.9836169,0.002427736,0.001361659,0.00276505,0.00007012396],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.174088,0.00001098773,0.8205755,0.003806435,0.00003341836,0.0002021512,1.467025e-7,0.00003077961,0.001252543],"genre_scores_gemma":[0.9784726,0.000004244024,0.021353,0.0000569663,0.00002822558,0.00004632606,0.00000105242,0.000003232505,0.00003436967],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9598567,"threshold_uncertainty_score":0.1197484,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2261500682115022,"score_gpt":0.3147457090188217,"score_spread":0.08859564080731949,"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."}}