{"id":"W4408925909","doi":"10.1145/3676641.3716009","title":"PAPI: Exploiting Dynamic Parallelism in Large Language Model Decoding with a Processing-In-Memory-Enabled Computing System","year":2025,"lang":"en","type":"article","venue":"","topic":"Parallel Computing and Optimization Techniques","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"Universitas Brawijaya","keywords":"Computer science; Parallel computing; Parallelism (grammar); Decoding methods; Computer architecture; Programming language; Algorithm","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":[0.0008614895,0.0002270807,0.0003471842,0.0005421513,0.000195678,0.000269156,0.000732393,0.0000914423,0.000001058355],"category_scores_gemma":[0.00003651464,0.0002011095,0.00003921038,0.001207822,0.00001818665,0.0004287747,0.000345403,0.0002806646,0.000002756206],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002368802,"about_ca_system_score_gemma":0.0001868802,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008317045,"about_ca_topic_score_gemma":0.0001474424,"domain_scores_codex":[0.9980261,0.000105225,0.0005016614,0.0005924374,0.0002104107,0.0005641786],"domain_scores_gemma":[0.9992383,0.00007555693,0.00015093,0.0004009131,0.00008239836,0.00005189192],"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":[0.00001740088,0.0001333871,0.003129544,0.000418725,0.000008839861,0.0001312833,0.005175909,0.9356161,0.0001720206,0.03635419,0.00006460645,0.01877797],"study_design_scores_gemma":[0.0008626571,0.00001551292,0.0004298242,0.001089448,0.000002706591,0.00001032274,0.001233608,0.9956719,0.000193166,0.0002498582,0.000002785077,0.0002382396],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06503608,0.0002283715,0.9179398,0.0001526466,0.00003928714,0.0002942739,3.866963e-7,0.001188914,0.01512025],"genre_scores_gemma":[0.6871401,0.000003718611,0.3122542,0.0002100917,0.00000623973,0.00001944295,0.000002023355,0.00001051577,0.0003536621],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.622104,"threshold_uncertainty_score":0.8201007,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00812880525198429,"score_gpt":0.2625035649182994,"score_spread":0.2543747596663151,"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."}}