{"id":"W3091097978","doi":"10.1145/3452296.3472904","title":"Efficient sparse collective communication and its application to accelerate distributed deep learning","year":2021,"lang":"en","type":"article","venue":"","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":105,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"China Scholarship Council; King Abdullah University of Science and Technology","keywords":"Computer science; Focus (optics); Distributed computing; Scale (ratio); Artificial intelligence","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":[],"consensus_categories":[],"category_scores_codex":[0.000160749,0.00008369364,0.0000945494,0.0000761513,0.0002245162,0.0001338069,0.0002849944,0.00003686758,0.00001262745],"category_scores_gemma":[0.000195363,0.00008767386,0.00001479152,0.0009472033,0.00001319125,0.00008277711,0.0003838631,0.00008718074,0.0000189203],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009155944,"about_ca_system_score_gemma":0.00004484086,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006007157,"about_ca_topic_score_gemma":0.00000692415,"domain_scores_codex":[0.9991888,0.00008406676,0.0001501532,0.0003054307,0.0001366807,0.0001348677],"domain_scores_gemma":[0.9990306,0.000108001,0.0000614565,0.0003451717,0.0003704251,0.00008429617],"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":[0.00001549518,0.0002396916,0.0004666215,0.00001678623,0.00003055967,0.000004087422,0.003890876,0.4713096,0.009848181,0.4895434,0.0004953446,0.02413942],"study_design_scores_gemma":[0.0001330144,0.00003221244,0.001407668,0.000009835304,0.000003391091,0.000007677604,0.00006992464,0.9793795,0.01756199,0.0009873527,0.000298539,0.00010889],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00885052,0.0001063689,0.9879081,0.001020357,0.00002262635,0.00034299,0.000001530816,0.0003345123,0.001412982],"genre_scores_gemma":[0.8609477,0.0000142908,0.1384776,0.0001894251,0.000004482165,0.0001174979,0.0000234431,0.000005690063,0.0002199313],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8520972,"threshold_uncertainty_score":0.3575236,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01822701639295126,"score_gpt":0.2550134356046615,"score_spread":0.2367864192117102,"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."}}