{"id":"W2114508814","doi":"","title":"Low precision arithmetic for deep learning","year":2014,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":88,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Polytechnique Montréal","funders":"","keywords":"MNIST database; Computer science; Benchmark (surveying); Computation; Artificial neural network; Point (geometry); Set (abstract data type); Floating point; Artificial intelligence; Training set; Deep learning; Algorithm; Machine learning; Arithmetic; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001657899,0.00101655,0.0006023642,0.0006626772,0.0005389798,0.001152692,0.001834265,0.001106879,0.01180808],"category_scores_gemma":[0.01123634,0.0003835846,0.0005564489,0.001291935,0.0008475677,0.002130756,0.0009878507,0.002559067,0.002119919],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001012034,"about_ca_system_score_gemma":0.001252291,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004794687,"about_ca_topic_score_gemma":0.004410666,"domain_scores_codex":[0.9991199,0.0002307685,0.00006764568,0.0001557508,0.0003111179,0.0001147865],"domain_scores_gemma":[0.9975836,0.001458036,0.0000995168,0.0004136683,0.0003995722,0.00004559793],"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":[0.001242096,0.000286343,0.003297634,0.00075053,0.0001314534,0.0002160638,0.0002128512,0.7915534,0.007375213,0.06949396,0.03192097,0.09351949],"study_design_scores_gemma":[0.0002320467,0.0002660872,0.0006850272,0.00009681313,0.00004151288,0.00009335927,0.00006459708,0.9028299,0.02146941,0.06306367,0.01112829,0.00002927215],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4002736,0.003939955,0.5285333,0.002199073,0.001146634,0.0003254825,0.006858893,0.01803005,0.03869303],"genre_scores_gemma":[0.7612242,0.0007212254,0.2269658,0.0005685491,0.00008229209,0.0005894714,0.00328026,0.001167906,0.005400271],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01180808,"threshold_uncertainty_score":0.03950191,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03463392661948356,"score_gpt":0.1742007141424249,"score_spread":0.1395667875229414,"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."}}