{"id":"https://openalex.org/W7128805589","doi":"https://doi.org/10.1145/3779212.3790180","title":"LAER-MoE: Load-Adaptive Expert Re-layout for Efficient Mixture-of-Experts Training","display_name":"LAER-MoE: Load-Adaptive Expert Re-layout for Efficient Mixture-of-Experts Training","publication_year":2026,"publication_date":"2026-03-10","ids":{"openalex":"https://openalex.org/W7128805589","doi":"https://doi.org/10.1145/3779212.3790180"},"language":null,"primary_location":{"id":"doi:10.1145/3779212.3790180","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3779212.3790180","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3779212.3790180","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5125927330","display_name":"Xinyi Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xinyi Liu","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0008-5863-2613","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125917429","display_name":"Yujie Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yujie Wang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0000-8375-493X","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125963862","display_name":"Fangcheng Fu","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fangcheng Fu","raw_affiliation_strings":["Shanghai Jiao Tong University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0003-1658-0380","affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Xuefeng Xiao","orcid":"https://orcid.org/0009-0009-8258-1243"},"institutions":[{"id":"https://openalex.org/I4210128111","display_name":"Clover Seed (China)","ror":"https://ror.org/037jwt041","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210128111"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuefeng Xiao","raw_affiliation_strings":["Bytedance Seed, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0009-8258-1243","affiliations":[{"raw_affiliation_string":"Bytedance Seed, Beijing, China","institution_ids":["https://openalex.org/I4210128111"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Huixia Li","orcid":"https://orcid.org/0009-0002-5968-3793"},"institutions":[{"id":"https://openalex.org/I4210128111","display_name":"Clover Seed (China)","ror":"https://ror.org/037jwt041","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210128111"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huixia Li","raw_affiliation_strings":["Bytedance Seed, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0002-5968-3793","affiliations":[{"raw_affiliation_string":"Bytedance Seed, Beijing, China","institution_ids":["https://openalex.org/I4210128111"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125946338","display_name":"Jiashi Li","orcid":null},"institutions":[{"id":"https://openalex.org/I4210098582","display_name":"ZTE (China)","ror":"https://ror.org/00rjhhq63","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210098582"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiashi Li","raw_affiliation_strings":["Bytedance Seed, Shenzhen, China"],"raw_orcid":"https://orcid.org/0009-0004-7171-8527","affiliations":[{"raw_affiliation_string":"Bytedance Seed, Shenzhen, China","institution_ids":["https://openalex.org/I4210098582"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5125931642","display_name":"Bin Cui","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Bin Cui","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1681-4677","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.16012774,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1055","last_page":"1072"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.6057999730110168,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.6057999730110168,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.03880000114440918,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.03280000016093254,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/expert-system","display_name":"Expert system","score":0.6790000200271606},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.5849999785423279},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.5539000034332275},{"id":"https://openalex.org/keywords/routing","display_name":"Routing (electronic design automation)","score":0.491100013256073},{"id":"https://openalex.org/keywords/load-balancing","display_name":"Load balancing (electrical power)","score":0.4480000138282776},{"id":"https://openalex.org/keywords/planner","display_name":"Planner","score":0.4408000111579895},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.36880001425743103},{"id":"https://openalex.org/keywords/granularity","display_name":"Granularity","score":0.3540000021457672}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7287999987602234},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.6790000200271606},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.5849999785423279},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.5539000034332275},{"id":"https://openalex.org/C74172769","wikidata":"https://www.wikidata.org/wiki/Q1446839","display_name":"Routing (electronic design automation)","level":2,"score":0.491100013256073},{"id":"https://openalex.org/C138959212","wikidata":"https://www.wikidata.org/wiki/Q1806783","display_name":"Load balancing (electrical power)","level":3,"score":0.4480000138282776},{"id":"https://openalex.org/C2776999362","wikidata":"https://www.wikidata.org/wiki/Q2349274","display_name":"Planner","level":2,"score":0.4408000111579895},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.3801000118255615},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.36880001425743103},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.3540000021457672},{"id":"https://openalex.org/C2779370713","wikidata":"https://www.wikidata.org/wiki/Q357554","display_name":"Load management","level":2,"score":0.3398999869823456},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.31369999051094055},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.3073999881744385},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.30480000376701355},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.2994000017642975},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.29179999232292175},{"id":"https://openalex.org/C19012869","wikidata":"https://www.wikidata.org/wiki/Q578372","display_name":"Response time","level":2,"score":0.2888000011444092},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.28839999437332153},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2874999940395355},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.27889999747276306},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2759000062942505},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.27300000190734863},{"id":"https://openalex.org/C2984173633","wikidata":"https://www.wikidata.org/wiki/Q22725","display_name":"Routing algorithm","level":4,"score":0.258899986743927},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.2547999918460846}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3779212.3790180","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3779212.3790180","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2602.11686","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2602.11686","pdf_url":"https://arxiv.org/pdf/2602.11686","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.1145/3779212.3790180","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3779212.3790180","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3648843532","display_name":null,"funder_award_id":"U23B2048","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320321106","display_name":"Ministry of Education of the People's Republic of China","ror":"https://ror.org/01mv9t934"},{"id":"https://openalex.org/F4320324787","display_name":"Peking University","ror":"https://ror.org/02v51f717"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Expert":[0,73],"parallelism":[1],"is":[2,66],"vital":[3],"for":[4,101,142],"effectively":[5],"training":[6,49,60,108,170],"Mixture-of-Experts":[7],"(MoE)":[8],"models,":[9],"enabling":[10],"different":[11,21],"devices":[12,86],"to":[13,109,122,133,162,166],"host":[14],"distinct":[15],"experts,":[16],"with":[17],"each":[18,79],"device":[19],"processing":[20],"input":[22],"data.":[23],"However,":[24],"during":[25,97,107,144],"expert":[26,80,92,105],"parallel":[27,69],"training,":[28],"dynamic":[29],"routing":[30,140],"results":[31,155],"in":[32],"significant":[33],"load":[34,111,130],"imbalance":[35],"among":[36],"experts:":[37],"a":[38,48,67,129],"handful":[39],"of":[40,64,85,104,119,137],"overloaded":[41],"experts":[42,90,138],"hinder":[43],"overall":[44],"iteration,":[45],"emerging":[46],"as":[47],"bottleneck.":[50],"In":[51,113],"this":[52],"paper,":[53],"we":[54,115,127],"introduce":[55],"LAER-MoE,":[56],"an":[57,150],"efficient":[58],"MoE":[59],"framework.":[61],"The":[62],"core":[63],"LAER-MoE":[65],"novel":[68],"paradigm,":[70],"Fully":[71],"Sharded":[72],"Parallel":[74],"(FSEP),":[75],"which":[76],"fully":[77],"partitions":[78],"parameter":[81],"by":[82],"the":[83,154,167],"number":[84],"and":[87,139,153],"restores":[88],"partial":[89],"at":[91,175],"granularity":[93],"through":[94],"All-to-All":[95],"communication":[96,120,124],"training.":[98,145],"This":[99],"allows":[100],"flexible":[102],"re-layout":[103,135],"parameters":[106],"enhance":[110],"balancing.":[112],"particular,":[114],"perform":[116,147],"fine-grained":[117],"scheduling":[118],"operations":[121],"minimize":[123],"overhead.":[125],"Additionally,":[126],"develop":[128],"balancing":[131],"planner":[132],"formulate":[134],"strategies":[136],"schemes":[141],"tokens":[143],"We":[146],"experiments":[148],"on":[149],"A100":[151],"cluster,":[152],"indicate":[156],"that":[157],"our":[158],"system":[159],"achieves":[160],"up":[161],"1.69x":[163],"acceleration":[164],"compared":[165],"current":[168],"state-of-the-art":[169],"systems.":[171],"Source":[172],"code":[173],"available":[174],"https://github.com/PKU-DAIR/Hetu-Galvatron/tree/laer-moe.":[176]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-02-14T00:00:00"}