{"id":"W3173842498","doi":"","title":"Low-Precision Training in Logarithmic Number System using Multiplicative Weight Update","year":2021,"lang":"en","type":"preprint","venue":"CaltechAUTHORS (California Institute of Technology)","topic":"Numerical Methods and Algorithms","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Semtech (Canada)","funders":"Nvidia; California Institute of Technology","keywords":"Computer science; Madam; Multiplicative function; Algorithm; Quantization (signal processing); Overhead (engineering); Convolutional neural network; Computer engineering; Artificial intelligence; Mathematics","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","research_integrity"],"consensus_categories":["research_integrity"],"category_scores_codex":[0.001270998,0.0007853459,0.001794484,0.001273918,0.0001950006,0.0001521757,0.003481997,0.001888329,0.00001278008],"category_scores_gemma":[0.0005914923,0.000757245,0.0004053485,0.00282731,0.0006465281,0.0003881086,0.005276398,0.002657604,0.00005015633],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006380327,"about_ca_system_score_gemma":0.0007851218,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001269967,"about_ca_topic_score_gemma":0.00001374833,"domain_scores_codex":[0.9945941,0.0002288445,0.00164687,0.00200993,0.0006009681,0.0009192738],"domain_scores_gemma":[0.9957935,0.00009569534,0.00105137,0.002422323,0.0004075755,0.0002295582],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003948311,0.0004623359,0.0006547438,0.00139373,0.000243003,0.001152573,0.0007393089,0.006868689,0.0147436,0.09363134,0.00003019598,0.880041],"study_design_scores_gemma":[0.001393431,0.00009044788,0.00008885618,0.007022177,0.0001057748,0.0004341929,0.0008560057,0.8199823,0.1053627,0.05412003,0.008538752,0.002005358],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.125856,0.0004147104,0.8683352,0.0008016202,0.002623473,0.0008484022,0.0001219699,0.0007951497,0.0002034091],"genre_scores_gemma":[0.3031985,0.0001054768,0.6963351,0.00004276347,0.00008570022,0.0001497247,0.00002865726,0.00004628345,0.00000780739],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8780357,"threshold_uncertainty_score":0.9996433,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03206347963469536,"score_gpt":0.304197529823886,"score_spread":0.2721340501891907,"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."}}