{"id":"W4200559323","doi":"10.1109/mlhpc54614.2021.00006","title":"Semantic-Aware Lossless Data Compression for Deep Learning Recommendation Model (DLRM)","year":2021,"lang":"en","type":"article","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Advanced Micro Devices (Canada)","funders":"Advanced Micro Devices","keywords":"Computer science; Deep learning; Bottleneck; Artificial neural network; Lossless compression; Deep neural networks; Inference; Artificial intelligence; Volume (thermodynamics); Data compression; Parallel computing; Machine learning; Embedded system","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":[0.0007239204,0.0008686681,0.0007284498,0.0006344136,0.0003773229,0.0007853133,0.001505054,0.0008233399,0.003769893],"category_scores_gemma":[0.002679446,0.0003246159,0.0006160355,0.001109842,0.0003743939,0.002159932,0.0009373052,0.001584717,0.001667547],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007875076,"about_ca_system_score_gemma":0.001192205,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009893224,"about_ca_topic_score_gemma":0.01619673,"domain_scores_codex":[0.9994074,0.00008509324,0.0000521095,0.0001168989,0.0002750542,0.00006345583],"domain_scores_gemma":[0.9993008,0.0001466538,0.00004895759,0.0002800038,0.0001956319,0.00002786791],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0004067854,0.0001937559,0.001651812,0.000168892,0.00009604536,0.0001703914,0.00008731662,0.1809325,0.02158862,0.009120981,0.01936027,0.7662227],"study_design_scores_gemma":[0.00002178985,0.00005395364,0.0002826414,0.000008756542,0.00001037671,0.00007454008,0.00001752925,0.9764082,0.0171681,0.00282473,0.003116466,0.0000129088],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03469183,0.000839468,0.9513596,0.0005231164,0.0001538325,0.00009708692,0.0008099953,0.008621258,0.002903863],"genre_scores_gemma":[0.3198974,0.0007072763,0.6665308,0.0005207167,0.0001036166,0.0002333122,0.003659958,0.0003628222,0.007984022],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009893224,"threshold_uncertainty_score":0.01967132,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09610207281267812,"score_gpt":0.3477033958897529,"score_spread":0.2516013230770748,"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."}}