{"id":"W4404133810","doi":"10.1145/3649329.3656516","title":"QUQ: Quadruplet Uniform Quantization for Efficient Vision Transformer Inference","year":2024,"lang":"en","type":"article","venue":"","topic":"CCD and CMOS Imaging Sensors","field":"Engineering","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"National Key Research and Development Program of China; National Natural Science Foundation of China","keywords":"Computer science; Transformer; Inference; Quantization (signal processing); Artificial intelligence; Computer vision; Speech recognition; Electrical engineering; Engineering","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":[0.00007989771,0.00009663626,0.00007725716,0.00008382693,0.00003452172,0.00008552551,0.00005108827,0.00003743145,0.00009344285],"category_scores_gemma":[0.000009127377,0.00007963623,0.00005364188,0.0001795582,0.0000122084,0.00009034088,0.000002476346,0.00006030136,0.00008580861],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003320731,"about_ca_system_score_gemma":0.00001189723,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001249539,"about_ca_topic_score_gemma":0.00001900641,"domain_scores_codex":[0.9994856,0.000002701322,0.0001367097,0.0001194202,0.00009070404,0.0001649198],"domain_scores_gemma":[0.9997918,0.00006424166,0.000003587693,0.00008429231,0.00002198126,0.00003407773],"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":[0.00001995766,0.00006128437,0.00008501585,0.001261646,0.00007526553,0.000005513021,0.002801895,0.6169868,0.05549477,0.131784,0.01170665,0.1797171],"study_design_scores_gemma":[0.0001093716,0.00002365129,0.00008473481,0.0000656772,0.00001116586,0.000001940179,0.00008263622,0.9417889,0.01186975,0.0002139875,0.04561859,0.0001296331],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.136427,0.0002617224,0.8506551,0.0003046295,0.0006849396,0.0002241011,0.00001607203,0.001016815,0.01040958],"genre_scores_gemma":[0.9980736,0.00004378574,0.001203288,0.00003018215,0.0000503282,0.00001155183,0.00002489517,0.00002688536,0.0005355153],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8616466,"threshold_uncertainty_score":0.3247471,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009876990074872903,"score_gpt":0.2698857335242779,"score_spread":0.260008743449405,"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."}}