{"id":"W2965174861","doi":"10.24963/ijcai.2019/525","title":"COP: Customized Deep Model Compression via Regularized Correlation-Based Filter-Level Pruning","year":2019,"lang":"en","type":"article","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":73,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institute for Advanced Research","keywords":"Computer science; Redundancy (engineering); Normalization (sociology); Convolutional neural network; Pruning; Filter (signal processing); Artificial neural network; Algorithm; Computational complexity theory; Artificial intelligence; Mathematical optimization; Mathematics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001595816,0.0001973335,0.0002522967,0.00009768656,0.0001976491,0.00007610214,0.000860065,0.0000964741,0.0001161023],"category_scores_gemma":[0.00002128972,0.000174294,0.00008456554,0.0004512649,0.00003555757,0.0005270002,0.0002436739,0.0002092593,0.0003726242],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006397469,"about_ca_system_score_gemma":0.00006454234,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006423237,"about_ca_topic_score_gemma":0.000004149764,"domain_scores_codex":[0.9983867,0.00007744533,0.0003348827,0.0005547563,0.0003265426,0.0003196151],"domain_scores_gemma":[0.9980831,0.0004078568,0.0001893863,0.001071548,0.0001242962,0.0001238401],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003672132,0.00004493416,0.0001951015,0.000006116203,0.000006099714,7.782843e-7,0.0000572074,0.9470533,0.0194261,0.02287336,0.0005389912,0.009761306],"study_design_scores_gemma":[0.001812966,0.00001867803,0.0003394528,0.0000229358,0.000005944122,0.000002708461,0.000002868346,0.9866821,0.002882504,0.00767711,0.0003201335,0.0002326462],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004209765,0.00002005785,0.9908051,0.0006473072,0.0001846433,0.0006896731,0.000002822519,0.0004938839,0.002946727],"genre_scores_gemma":[0.4960853,0.000001121984,0.5009042,0.0006298512,0.0000160276,0.0000474288,0.0000169925,0.00001418749,0.002284901],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.4918755,"threshold_uncertainty_score":0.7107503,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02319882392590332,"score_gpt":0.2510254103924932,"score_spread":0.2278265864665899,"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."}}