{"id":"W3093046553","doi":"10.1145/3466752.3480106","title":"FPRaker: A Processing Element For Accelerating Neural Network Training","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; Toronto Public Health; University of Toronto","funders":"","keywords":"Accumulator (cryptography); Computer science; Operand; Floating point; Quantization (signal processing); Artificial neural network; Training (meteorology); Pruning; Algorithm; Artificial intelligence; Computer hardware","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004903729,0.0009460281,0.0004553815,0.0005922128,0.0005125734,0.0009690878,0.002169756,0.0007533808,0.01655843],"category_scores_gemma":[0.001978568,0.0005468743,0.0004403457,0.0005937577,0.0004023437,0.00149241,0.0009430081,0.001673176,0.004879639],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006802181,"about_ca_system_score_gemma":0.0009672416,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001526703,"about_ca_topic_score_gemma":0.003265499,"domain_scores_codex":[0.9995667,0.00004633306,0.00002688563,0.00009566042,0.0002102929,0.00005407417],"domain_scores_gemma":[0.9994374,0.0001982616,0.00005033438,0.0001018203,0.0001679841,0.00004424165],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00117728,0.0002970923,0.002639829,0.0006764133,0.0002212472,0.0005533744,0.0003699097,0.09547958,0.1634642,0.02747736,0.05632296,0.6513208],"study_design_scores_gemma":[0.0001078409,0.0004483053,0.0009405952,0.00007885885,0.00007231763,0.0004233962,0.00005750014,0.7356514,0.1836392,0.01162635,0.06687642,0.00007787815],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02379658,0.000723825,0.9376829,0.0002911849,0.0004088149,0.0001305633,0.0003973118,0.02479549,0.01177344],"genre_scores_gemma":[0.2032984,0.0004261124,0.7708541,0.0005148796,0.0001447524,0.0003214803,0.001171566,0.002023822,0.0212449],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01655843,"threshold_uncertainty_score":0.05539352,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1045721705117997,"score_gpt":0.317957734257139,"score_spread":0.2133855637453392,"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."}}