{"id":"W3087194612","doi":"10.48550/arxiv.2009.08576","title":"Pruning Neural Networks at Initialization: Why are We Missing the Mark?","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":48,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Initialization; Pruning; Computer science; Artificial neural network; Fraction (chemistry); Shuffling; Heuristics; Machine learning; Weighting; Artificial intelligence; Baseline (sea); Layer (electronics); GRASP","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":[0.0001132176,0.0003617399,0.0003003496,0.00008614337,0.0007688972,0.0002319435,0.002598943,0.0002130703,0.00002263331],"category_scores_gemma":[0.00002981681,0.0003555576,0.0001826384,0.001225038,0.0001721632,0.0004037147,0.004523378,0.0008464227,0.00002569432],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002184713,"about_ca_system_score_gemma":0.00006052616,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001118394,"about_ca_topic_score_gemma":0.00002651527,"domain_scores_codex":[0.997758,0.0002481938,0.0002487818,0.001209043,0.0001229027,0.0004130253],"domain_scores_gemma":[0.9974119,0.0002864035,0.0005083166,0.001454801,0.0001310644,0.0002075064],"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":[0.00001449917,0.00001452987,0.0006868617,0.00002419674,0.00002878885,0.0001518631,0.0001653258,0.9524915,0.000007998796,0.04188927,0.003242019,0.001283186],"study_design_scores_gemma":[0.0001631252,0.00001271313,0.0003211785,0.00008056535,0.0000438816,0.00001269539,0.00003792997,0.9582433,0.0000156289,0.0330078,0.00770448,0.0003567061],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002861408,0.0003130097,0.9845075,0.00994396,0.0005733469,0.0004653272,0.000006072564,0.0004742602,0.0008550887],"genre_scores_gemma":[0.9941382,0.000457305,0.002431319,0.00212275,0.0003775714,0.000004291997,0.0000280589,0.00003502949,0.0004054221],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9912769,"threshold_uncertainty_score":0.9998896,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1203842908523943,"score_gpt":0.2164573225010282,"score_spread":0.09607303164863397,"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."}}