{"id":"https://openalex.org/W7134845083","doi":"https://doi.org/10.48550/arxiv.2603.08058","title":"Stabilized Fine-Tuning with LoRA in Federated Learning: Mitigating the Side Effect of Client Size and Rank via the Scaling Factor","display_name":"Stabilized Fine-Tuning with LoRA in Federated Learning: Mitigating the Side Effect of Client Size and Rank via the Scaling Factor","publication_year":2026,"publication_date":"2026-03-09","ids":{"openalex":"https://openalex.org/W7134845083","doi":"https://doi.org/10.48550/arxiv.2603.08058"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2603.08058","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"type":"article","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5106976370","display_name":"Jiayu Huang","orcid":"https://orcid.org/0009-0003-8964-4696"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Jiayu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128637063","display_name":"Xiaohu Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Xiaohu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128681931","display_name":"Tiantian He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Tiantian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5013010213","display_name":"Qicheng Lao","orcid":"https://orcid.org/0000-0002-6032-8548"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lao, Qicheng","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":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.29159012,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.42980000376701355,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.42980000376701355,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.19820000231266022,"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.03629999980330467,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.5889999866485596},{"id":"https://openalex.org/keywords/rank","display_name":"Rank (graph theory)","score":0.5205000042915344},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.49900001287460327},{"id":"https://openalex.org/keywords/factor","display_name":"Factor (programming language)","score":0.4950000047683716},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.47929999232292175},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.47110000252723694},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.4672999978065491},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.45719999074935913}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8399999737739563},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.5889999866485596},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.5205000042915344},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.49900001287460327},{"id":"https://openalex.org/C2781039887","wikidata":"https://www.wikidata.org/wiki/Q1391724","display_name":"Factor (programming language)","level":2,"score":0.4950000047683716},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.47929999232292175},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.47110000252723694},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.4672999978065491},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.45719999074935913},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4505000114440918},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.4269999861717224},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3449000120162964},{"id":"https://openalex.org/C2992525071","wikidata":"https://www.wikidata.org/wiki/Q50818671","display_name":"Federated learning","level":2,"score":0.3384000062942505},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3206999897956848},{"id":"https://openalex.org/C86037889","wikidata":"https://www.wikidata.org/wiki/Q4330127","display_name":"Learning to rank","level":3,"score":0.32030001282691956},{"id":"https://openalex.org/C23130292","wikidata":"https://www.wikidata.org/wiki/Q5275358","display_name":"Differential privacy","level":2,"score":0.3003000020980835},{"id":"https://openalex.org/C2776434776","wikidata":"https://www.wikidata.org/wiki/Q19246213","display_name":"Domain adaptation","level":3,"score":0.29910001158714294},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.29440000653266907},{"id":"https://openalex.org/C57869625","wikidata":"https://www.wikidata.org/wiki/Q1783502","display_name":"Rate of convergence","level":3,"score":0.29409998655319214},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2939000129699707},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.2930999994277954},{"id":"https://openalex.org/C2777131613","wikidata":"https://www.wikidata.org/wiki/Q18394147","display_name":"Retard","level":2,"score":0.28929999470710754},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.26179999113082886},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.26030001044273376},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2581999897956848},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.2578999996185303}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2603.08058","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2603.08058","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.08058","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:doi:10.48550/arxiv.2603.08058","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},"sustainable_development_goals":[{"score":0.462492972612381,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4,172],"pivotal":[5],"in":[6,141,163],"natural":[7],"language":[8],"processing.":[9],"The":[10],"impracticality":[11],"of":[12,44,148],"full":[13],"fine-tuning":[14],"has":[15],"prompted":[16],"Parameter-Efficient":[17],"Fine-Tuning":[18],"(PEFT)":[19],"methods":[20],"like":[21],"Low-Rank":[22],"Adaptation":[23],"(LoRA),":[24],"optimizing":[25],"low-rank":[26],"matrices":[27],"A":[28],"and":[29,115,168,185,190],"B.":[30],"In":[31],"distributed":[32],"scenarios":[33],"where":[34],"privacy":[35],"constraints":[36],"necessitate":[37],"Federated":[38,102],"Learning":[39],"(FL),":[40],"however,":[41],"the":[42,64,80,87,91,110,128,137,146,153],"integration":[43],"LoRA":[45,103],"is":[46],"often":[47],"unstable.":[48],"Specifically,":[49],"we":[50],"identify":[51],"that":[52,61,107,180],"aggregating":[53],"updates":[54],"from":[55],"multiple":[56],"clients":[57],"introduces":[58,100],"statistical":[59],"variance":[60],"scales":[62],"with":[63,194],"client":[65],"count,":[66],"causing":[67],"gradient":[68],"collapse":[69],"when":[70],"using":[71],"high-rank":[72,149,183,198],"adapters.":[73],"Existing":[74],"scaling":[75,122,138],"factor":[76,123],"candidates,":[77],"such":[78],"as":[79],"one":[81],"used":[82],"by":[83,90],"Rank-Stabilized":[84],"LoRA,":[85],"ignore":[86],"interaction":[88,111],"caused":[89],"aggregation":[92,129],"process.":[93],"To":[94],"bridge":[95],"this":[96,98],"gap,":[97],"paper":[99],"Stabilized":[101],"(SFed-LoRA),":[104],"a":[105],"framework":[106],"theoretically":[108],"characterizes":[109],"between":[112],"adapter":[113],"rank":[114],"federated":[116],"aggregation.":[117],"We":[118,178],"derive":[119],"an":[120],"optimal":[121],"designed":[124],"to":[125,174],"effectively":[126],"mitigate":[127],"error":[130],"accumulating":[131],"across":[132],"N":[133],"clients.":[134],"By":[135],"correcting":[136],"mismatch":[139],"inherent":[140],"previous":[142],"approaches,":[143],"SFed-LoRA":[144,181],"restores":[145],"efficacy":[147],"adaptation":[150],"without":[151],"altering":[152],"original":[154],"model":[155,166],"architecture":[156],"or":[157],"increasing":[158],"inference":[159],"latency.":[160],"Extensive":[161],"experiments":[162],"diverse":[164],"tasks,":[165],"architectures,":[167],"heterogeneous":[169],"data":[170],"distributions":[171],"conducted":[173],"validate":[175],"our":[176],"results.":[177],"demonstrate":[179],"prevents":[182],"collapse,":[184],"achieves":[186],"significantly":[187],"improved":[188],"stability":[189],"faster":[191],"convergence":[192],"compared":[193],"state-of-the-art":[195],"baselines":[196],"for":[197],"adaptation.":[199]},"counts_by_year":[],"updated_date":"2026-07-15T18:14:33.161393","created_date":"2026-03-11T00:00:00"}