{"id":"W3096344818","doi":"10.1109/cvpr46437.2021.01544","title":"Permute, Quantize, and Fine-tune: Efficient Compression of Neural Networks","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; University of Waterloo","funders":"","keywords":"Vector quantization; Computer science; Quantization (signal processing); Algorithm; Uncompressed video; Image compression; Artificial neural network; Data compression ratio; Convolutional neural network; Theoretical computer science; Artificial intelligence; Image processing; Image (mathematics); Object (grammar)","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":[0.0001202657,0.0002967914,0.0004567296,0.00008266063,0.0001183577,0.0001431141,0.001045827,0.0001964141,0.00001564301],"category_scores_gemma":[0.0000209249,0.0002620123,0.0001236557,0.0003822454,0.0001106572,0.0001027569,0.004402793,0.0005862407,0.000001271733],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000200968,"about_ca_system_score_gemma":0.00004479864,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002997598,"about_ca_topic_score_gemma":0.00001779103,"domain_scores_codex":[0.9979053,0.00009742565,0.0004813964,0.000911967,0.0002974508,0.0003064998],"domain_scores_gemma":[0.9977531,0.0002568818,0.0003159114,0.001367363,0.0001724179,0.0001342686],"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":[0.000004754837,0.00008482041,0.0002370236,0.00006309085,0.0000148918,0.000009788432,0.00009406653,0.9661004,0.0006137277,0.0113705,0.0005984259,0.02080854],"study_design_scores_gemma":[0.0001628333,0.00001933196,0.002517065,0.0001069083,0.00001135531,0.00001844799,0.00001057791,0.9955479,0.0007048465,0.0004931627,0.000155862,0.0002516993],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09988536,0.002339592,0.8955266,0.0006397354,0.0005602914,0.0004108447,0.000003946581,0.0001897923,0.0004438676],"genre_scores_gemma":[0.9197692,0.0002031059,0.07954799,0.0001553078,0.00008891484,0.00005042417,0.00003187834,0.00001757952,0.0001356073],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8198838,"threshold_uncertainty_score":0.9999832,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02202910622663319,"score_gpt":0.2797244536248866,"score_spread":0.2576953473982534,"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."}}