{"id":"W4402670576","doi":"10.18653/v1/2024.findings-acl.621","title":"Mixture-of-Supernets: Improving Weight-Sharing Supernet Training with Architecture-Routed Mixture-of-Experts","year":2024,"lang":"en","type":"article","venue":"","topic":"Online Learning and Analytics","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Alliance de recherche numérique du Canada; Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs","keywords":"Computer science; Training (meteorology); Architecture; Computer architecture; Multimedia","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003326148,0.002365643,0.003025856,0.00131082,0.001093465,0.001159466,0.004824323,0.003515973,0.006245744],"category_scores_gemma":[0.008472703,0.001477922,0.002325654,0.001625445,0.0009655659,0.003672652,0.003466989,0.004621709,0.003217512],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000886352,"about_ca_system_score_gemma":0.001968342,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01084505,"about_ca_topic_score_gemma":0.01853637,"domain_scores_codex":[0.9985195,0.0006496767,0.00007547008,0.0003550912,0.0001949614,0.0002053136],"domain_scores_gemma":[0.9967958,0.001948235,0.0001094728,0.0004529224,0.0004661755,0.0002274414],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001543952,0.0006562651,0.002094434,0.0001939923,0.0005889601,0.0002287421,0.0002214366,0.4810255,0.00349195,0.006853573,0.02179215,0.4813091],"study_design_scores_gemma":[0.00004244,0.00007685104,0.00009988493,0.0000149508,0.00003568926,0.00002904612,0.0000231533,0.9944963,0.0006433619,0.003849061,0.000680004,0.000009352245],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04910605,0.003211109,0.9355558,0.0007255536,0.0006390187,0.0001505086,0.0003989072,0.005428215,0.004784801],"genre_scores_gemma":[0.5738115,0.001170886,0.4057089,0.001769765,0.0005076113,0.0003662727,0.003201042,0.001138074,0.01232584],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01084505,"threshold_uncertainty_score":0.02156389,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01260724651315749,"score_gpt":0.2406474297289617,"score_spread":0.2280401832158042,"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."}}