{"id":"W4414405631","doi":"10.1109/clusterworkshops65972.2025.11164211","title":"A similarity-aware MOE-based method for optimizing tensor programs across diverse GPUs","year":2025,"lang":"en","type":"article","venue":"","topic":"Parallel Computing and Optimization Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Youth Innovation Promotion Association; Chinese Academy of Sciences; National Natural Science Foundation of China","keywords":"Representation (politics); Compiler; Tensor (intrinsic definition); Multi-core processor; Artificial neural network; Core (optical fiber)","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":[0.0006767142,0.0001892182,0.0002439654,0.0001223089,0.0004222647,0.000407688,0.0009505593,0.0001199026,0.000004100928],"category_scores_gemma":[0.00009319644,0.0001699311,0.0001607337,0.0006246777,0.00003972958,0.0002153104,0.0003690197,0.0001306769,0.000003269358],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005588596,"about_ca_system_score_gemma":0.0001149196,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005811888,"about_ca_topic_score_gemma":0.00001058682,"domain_scores_codex":[0.9984816,0.0000912478,0.0002714818,0.0005494591,0.0001655833,0.0004406341],"domain_scores_gemma":[0.9986799,0.0002736941,0.0000891073,0.0005725135,0.0003067843,0.00007797024],"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":[0.0001003639,0.0006937364,0.001746073,0.0002868359,0.0001392487,0.00002023312,0.001130678,0.2562786,0.0002016934,0.06116153,0.02418671,0.6540543],"study_design_scores_gemma":[0.0006449761,0.00009022081,0.00003940578,0.00005120876,0.000009187333,0.00000140038,0.00006341337,0.9847175,0.005689749,0.001906755,0.006576966,0.0002092077],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0001169138,0.00003064009,0.9940951,0.001791843,0.0001969801,0.0006450472,0.000005325322,0.001793051,0.001325098],"genre_scores_gemma":[0.01734141,0.000003047624,0.9789753,0.002023174,0.00002206382,0.0001339842,0.00001158252,0.000009849005,0.001479566],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7284389,"threshold_uncertainty_score":0.6929588,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03564461568228783,"score_gpt":0.3559284757428017,"score_spread":0.3202838600605139,"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."}}