{"id":"https://openalex.org/W7160893354","doi":"https://doi.org/10.48550/arxiv.2605.08738","title":"SlimQwen: Exploring the Pruning and Distillation in Large MoE Model Pre-training","display_name":"SlimQwen: Exploring the Pruning and Distillation in Large MoE Model Pre-training","publication_year":2026,"publication_date":"2026-05-09","ids":{"openalex":"https://openalex.org/W7160893354","doi":"https://doi.org/10.48550/arxiv.2605.08738"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.08738","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08738","pdf_url":null,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.08738","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135910366","display_name":"Shengkun Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Shengkun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135964746","display_name":"Zekun Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zekun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135970675","display_name":"Bo Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Bo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135985633","display_name":"Liangyu Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Liangyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004626105","display_name":"Rui Men","orcid":"https://orcid.org/0000-0002-4429-3461"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Men, Rui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135985952","display_name":"Siqi Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Siqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135935293","display_name":"Xiulong Yuan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuan, Xiulong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135993662","display_name":"Zihan Qiu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiu, Zihan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135959414","display_name":"Zhiqiang Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Zhiqiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135914911","display_name":"Dayiheng Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Dayiheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.20280000567436218,"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"}},"topics":[{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.20280000567436218,"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"}},{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.17579999566078186,"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"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.10559999942779541,"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/pruning","display_name":"Pruning","score":0.8615999817848206},{"id":"https://openalex.org/keywords/initialization","display_name":"Initialization","score":0.7803000211715698},{"id":"https://openalex.org/keywords/compression","display_name":"Compression (physics)","score":0.5602999925613403},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.5049999952316284},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.44029998779296875},{"id":"https://openalex.org/keywords/distillation","display_name":"Distillation","score":0.42149999737739563},{"id":"https://openalex.org/keywords/simple","display_name":"Simple (philosophy)","score":0.41190001368522644}],"concepts":[{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.8615999817848206},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.7803000211715698},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7458999752998352},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6394000053405762},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5697000026702881},{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.5602999925613403},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5049999952316284},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.44029998779296875},{"id":"https://openalex.org/C204030448","wikidata":"https://www.wikidata.org/wiki/Q101017","display_name":"Distillation","level":2,"score":0.42149999737739563},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.41190001368522644},{"id":"https://openalex.org/C2781235140","wikidata":"https://www.wikidata.org/wiki/Q275131","display_name":"Scratch","level":2,"score":0.34450000524520874},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.31619998812675476},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.3093000054359436},{"id":"https://openalex.org/C25797200","wikidata":"https://www.wikidata.org/wiki/Q828137","display_name":"Compression ratio","level":3,"score":0.2946000099182129},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.27959999442100525},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.26899999380111694},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.26759999990463257},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.25029999017715454}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.08738","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08738","pdf_url":null,"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.08738","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.08738","pdf_url":null,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Structured":[0],"pruning":[1,49,88,171],"and":[2,69,85],"knowledge":[3],"distillation":[4],"(KD)":[5],"are":[6],"typical":[7],"techniques":[8],"for":[9,206],"compressing":[10],"large":[11],"language":[12,143],"models,":[13],"but":[14],"it":[15,187],"remains":[16],"unclear":[17],"how":[18,58],"they":[19],"should":[20],"be":[21],"applied":[22],"at":[23,210],"pretraining":[24],"scale,":[25],"especially":[26],"to":[27,112,182,193],"recent":[28],"mixture-of-experts":[29],"(MoE)":[30],"models.":[31],"In":[32],"this":[33],"work,":[34],"we":[35,123,190],"systematically":[36],"study":[37],"MoE":[38,91,208],"compression":[39,60,109,209],"in":[40],"large-scale":[41,117],"pretraining,":[42],"focusing":[43],"on":[44,150],"three":[45],"key":[46],"questions:":[47],"whether":[48],"provides":[50],"a":[51,89,125,194],"better":[52,183],"initialization":[53],"than":[54],"training":[55,71,94,103,168],"from":[56,98],"scratch,":[57],"expert":[59,86,108,128],"choices":[61],"affect":[62],"the":[63,78,95,101,142,166],"final":[64,114],"model":[65,196],"after":[66,116],"continued":[67],"training,":[68],"which":[70,160],"strategy":[72,130],"is":[73],"most":[74,136],"effective.":[75],"We":[76,153],"have":[77],"following":[79],"findings:":[80],"First,":[81],"across":[82,135],"depth,":[83],"width,":[84],"compression,":[87,175],"pretrained":[90],"consistently":[92],"outperforms":[93,146],"target":[96],"architecture":[97,179],"scratch":[99],"under":[100],"same":[102,167],"budget.":[104],"Second,":[105],"different":[106],"one-shot":[107,174],"methods":[110],"converge":[111],"similar":[113],"performance":[115,134],"continual":[118],"pretraining.":[119],"Motivated":[120],"by":[121],"this,":[122],"introduce":[124],"simple":[126],"partial-preservation":[127],"merging":[129],"that":[131,177,197],"improves":[132],"downstream":[133],"benchmarks.":[137],"Third,":[138],"combining":[139],"KD":[140,147],"with":[141],"modeling":[144],"loss":[145],"alone,":[148],"particularly":[149],"knowledge-intensive":[151],"tasks.":[152],"further":[154],"propose":[155],"multi-token":[156],"prediction":[157],"(MTP)":[158],"distillation,":[159],"yields":[161],"consistent":[162],"gains.":[163],"Finally,":[164],"given":[165],"tokens,":[169],"progressive":[170],"schedules":[172],"outperform":[173],"suggesting":[176],"gradual":[178],"transitions":[180],"lead":[181],"optimization":[184],"trajectories.":[185],"Putting":[186],"all":[188],"together,":[189],"compress":[191],"Qwen3-Next-80A3B":[192],"23A2B":[195],"retains":[198],"competitive":[199],"performance.":[200],"These":[201],"results":[202],"offer":[203],"practical":[204],"guidance":[205],"efficient":[207],"scale.":[211]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-13T00:00:00"}