{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008093336,0.0005866077,0.0005678575,0.0004872961,0.0004177867,0.001056051,0.001517092,0.0005814681,0.004985133],"category_scores_gemma":[0.002749185,0.0003356433,0.0003980345,0.000702796,0.0005751988,0.001902525,0.0015054,0.001100851,0.001001909],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00063325,"about_ca_system_score_gemma":0.001110419,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003342138,"about_ca_topic_score_gemma":0.004997725,"domain_scores_codex":[0.9994838,0.00008162512,0.00003834057,0.0001372596,0.0002002786,0.00005864369],"domain_scores_gemma":[0.9994783,0.0001506472,0.00004480723,0.000141673,0.0001491657,0.00003544655],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006426654,0.000149129,0.001318646,0.0003036328,0.00006175683,0.0001614885,0.000213472,0.1004757,0.07294476,0.05097735,0.01724591,0.7555056],"study_design_scores_gemma":[0.00003421712,0.0001055143,0.0002544143,0.00001784736,0.00001352076,0.0001195452,0.00003751463,0.9456495,0.03083941,0.01743442,0.005469589,0.00002446673],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005852281,0.0002427881,0.9907711,0.00009996325,0.00005707506,0.0000442926,0.0001433833,0.001798809,0.0009903704],"genre_scores_gemma":[0.3799762,0.0003599789,0.6149113,0.0004075368,0.00006320903,0.0001419901,0.0008942226,0.0002791753,0.002966401],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004985133,"threshold_uncertainty_score":0.0166769,"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."}}