{"id":"W3152607317","doi":"10.18653/v1/2021.eacl-main.212","title":"Annealing Knowledge Distillation","year":2021,"lang":"en","type":"article","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":69,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Distillation; Artificial neural network; Benchmark (surveying); Simulated annealing; Inference; Artificial intelligence; Annealing (glass); Machine learning; Deep learning; Exploit; Materials science","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":[0.00002642264,0.00003771373,0.00003873727,0.00001252799,0.00007549205,0.00003943206,0.0001630246,0.00001343922,0.00001887592],"category_scores_gemma":[0.00001514974,0.00003598074,0.00001763348,0.0003959435,0.000007562499,0.0001785641,0.0001220126,0.0000336021,0.0001305638],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001304825,"about_ca_system_score_gemma":0.00002161931,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":4.953807e-7,"about_ca_topic_score_gemma":0.000007310867,"domain_scores_codex":[0.9995916,0.00001361349,0.0000740652,0.0001825111,0.00004694351,0.00009122966],"domain_scores_gemma":[0.9995177,0.0000616257,0.00001699652,0.0003006688,0.00006714524,0.00003582718],"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":[1.834908e-7,0.00001721815,0.0002262678,0.000002133871,0.000001907122,0.00000337984,0.00007619933,0.001612402,0.00225929,0.8670021,0.001645383,0.1271535],"study_design_scores_gemma":[0.0001526405,0.000009437766,0.006014573,0.000008550737,0.000002525672,0.00003470889,0.0000166892,0.7309002,0.0274514,0.06309384,0.1720942,0.0002212034],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00192359,0.0001888218,0.9614235,0.001087879,0.00009621499,0.00003126522,2.270147e-7,0.0001867296,0.03506184],"genre_scores_gemma":[0.7951962,0.0000155547,0.2020358,0.0001701156,0.00007398392,0.000007992047,0.000003638546,0.000003346941,0.002493301],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8039083,"threshold_uncertainty_score":0.1678177,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02531000715964295,"score_gpt":0.2946477081653764,"score_spread":0.2693377010057335,"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."}}