{"id":"W2612864759","doi":"10.1145/3061639.3062259","title":"Hardware-Software Codesign of Accurate, Multiplier-free Deep Neural Networks","year":2017,"lang":"en","type":"article","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":66,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institute for Advanced Research; Nvidia; National Science Foundation","keywords":"Computer science; Floating point; Multiplier (economics); Inference; Artificial neural network; Adder; Latency (audio); Computer engineering; Deep neural networks; Binary number; Overhead (engineering); Computer hardware; Hardware acceleration; Deep learning; Algorithm; Field-programmable gate array; Artificial intelligence; Arithmetic","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":[0.0001280959,0.0001977704,0.0002445113,0.00004457869,0.0005596816,0.0002069049,0.004530083,0.00008427042,0.00002537186],"category_scores_gemma":[0.0002589772,0.0001751143,0.00009316106,0.0001777547,0.0001692184,0.001051999,0.001328623,0.0001938851,0.00002001423],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002155572,"about_ca_system_score_gemma":0.00001806908,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002144172,"about_ca_topic_score_gemma":0.00005610291,"domain_scores_codex":[0.998512,0.00003787338,0.0003278494,0.0004999649,0.0002302234,0.0003921339],"domain_scores_gemma":[0.9955147,0.0002916896,0.000373193,0.003512099,0.0001633212,0.000145008],"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":[0.00004347913,0.0001970747,0.007473137,0.0000320255,0.00005388218,0.0000334688,0.0001981465,0.4446677,0.0009120894,0.098843,0.01187206,0.435674],"study_design_scores_gemma":[0.000448816,0.00004401281,0.007900085,0.000007783882,0.000005182661,0.000009874213,0.000004766511,0.9831012,0.001162554,0.005809398,0.0012765,0.0002297786],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001001888,0.00009858453,0.9956161,0.000840474,0.0003063247,0.0003340093,0.000004665025,0.0003295234,0.001468413],"genre_scores_gemma":[0.6985894,0.00002069981,0.3004586,0.0002595695,0.0000967915,0.00005100815,0.000002971698,0.00001597959,0.0005049205],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6975875,"threshold_uncertainty_score":0.8418097,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0307591957431979,"score_gpt":0.2835629074287377,"score_spread":0.2528037116855398,"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."}}