{"id":"W2997709794","doi":"10.1609/aaai.v34i04.5793","title":"On the Discrepancy between the Theoretical Analysis and Practical Implementations of Compressed Communication for Distributed Deep Learning","year":2020,"lang":"en","type":"article","venue":"","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Computer science; Implementation; Quantization (signal processing); Compression (physics); Convergence (economics); Compression ratio; Rate of convergence; Data compression; Bounded function; Upper and lower bounds; Algorithm; Data compression ratio; Layer (electronics); Theoretical computer science; Artificial intelligence; Image compression; Mathematics; Telecommunications; Image processing; Engineering; Channel (broadcasting)","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":[0.01574895,0.002826643,0.002247809,0.001940788,0.001723994,0.005618819,0.004718577,0.004782404,0.008091575],"category_scores_gemma":[0.1110867,0.001488319,0.001050881,0.00243326,0.007399196,0.01762972,0.006287971,0.01412249,0.002052861],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003017136,"about_ca_system_score_gemma":0.003196795,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001801168,"about_ca_topic_score_gemma":0.001741531,"domain_scores_codex":[0.987228,0.004985816,0.0006834485,0.00168795,0.004680285,0.0007345384],"domain_scores_gemma":[0.9036753,0.07658472,0.001983909,0.01087364,0.006113595,0.0007688852],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004787379,0.0002814427,0.001227983,0.001062643,0.00008487259,0.0002130314,0.0004495165,0.152606,0.003853462,0.700158,0.01444736,0.1251371],"study_design_scores_gemma":[0.00008989352,0.0002009354,0.0004758384,0.0006082924,0.0000334508,0.0003370827,0.0001994706,0.6279832,0.005707804,0.3560908,0.008199372,0.00007381194],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01954604,0.0148209,0.9260584,0.01555525,0.0007004965,0.0001223635,0.0002192926,0.0009618105,0.02201554],"genre_scores_gemma":[0.647765,0.02236801,0.3117165,0.005531863,0.002841788,0.001039646,0.0005544372,0.001239621,0.006943154],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01574895,"threshold_uncertainty_score":0.08328938,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03833580488631928,"score_gpt":0.3304518789392912,"score_spread":0.2921160740529719,"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."}}