{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003839808,0.0006329525,0.000545845,0.0005844435,0.0005326899,0.0005636136,0.001395532,0.000669658,0.002822389],"category_scores_gemma":[0.001485537,0.0004064107,0.0006555167,0.0006277849,0.0004997398,0.001059944,0.0007896,0.0009739571,0.0005690175],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007745668,"about_ca_system_score_gemma":0.001553279,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007248723,"about_ca_topic_score_gemma":0.01262025,"domain_scores_codex":[0.99976,0.00003747733,0.00001019554,0.0000465257,0.000105473,0.00004026019],"domain_scores_gemma":[0.9996164,0.0001004701,0.00003790021,0.00008460049,0.0001236254,0.00003706338],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008917847,0.00008340672,0.001219192,0.00006179174,0.00005063681,0.00005477087,0.00006752331,0.8539217,0.008902704,0.01180702,0.002898009,0.120844],"study_design_scores_gemma":[0.000003808394,0.000008698982,0.00003827234,0.000001188397,0.000002039179,0.000006021777,0.000004039338,0.9976959,0.0006505315,0.001193955,0.0003934269,0.000001974499],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02737199,0.0001534541,0.9658353,0.0002105042,0.00004421413,0.00004208823,0.00006088624,0.002879852,0.003401734],"genre_scores_gemma":[0.3866814,0.0001199982,0.6066989,0.0001769298,0.00004203013,0.0001710343,0.0002696659,0.0009483721,0.004891718],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007248723,"threshold_uncertainty_score":0.01441306,"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."}}