{"id":"https://openalex.org/W7130717034","doi":"https://doi.org/10.48550/arxiv.2602.17095","title":"FLoRG: Federated Fine-tuning with Low-rank Gram Matrices and Procrustes Alignment","display_name":"FLoRG: Federated Fine-tuning with Low-rank Gram Matrices and Procrustes Alignment","publication_year":2026,"publication_date":"2026-02-19","ids":{"openalex":"https://openalex.org/W7130717034","doi":"https://doi.org/10.48550/arxiv.2602.17095"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2602.17095","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2602.17095","pdf_url":"https://arxiv.org/pdf/2602.17095","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,"listed_in":[],"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"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2602.17095","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5126519825","display_name":"Chuiyang Meng","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Meng, Chuiyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128247290","display_name":"Ming Tang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I3045169105","display_name":"Southern University of Science and Technology","ror":"https://ror.org/049tv2d57","country_code":"CN","type":"education","lineage":["https://openalex.org/I3045169105"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Tang, Ming","raw_affiliation_strings":["Southern University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southern University of Science and Technology","institution_ids":["https://openalex.org/I3045169105"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5003037377","display_name":"Vincent W. S. Wong","orcid":"https://orcid.org/0000-0003-3821-4365","observed_orcids":["https://orcid.org/0000-0003-3821-4365"]},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Wong, Vincent W. S.","raw_affiliation_strings":["The University of British Columbia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of British Columbia","institution_ids":["https://openalex.org/I141945490"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":2,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9878000020980835,"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":0.9878000020980835,"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":0.007799999788403511,"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":0.001500000013038516,"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/federated-learning","display_name":"federated learning","score":0.8859999775886536},{"id":"https://openalex.org/keywords/low-rank-adaptation","display_name":"low-rank adaptation","score":0.7900000214576721},{"id":"https://openalex.org/keywords/lora","display_name":"LoRa","score":0.6940000057220459},{"id":"https://openalex.org/keywords/procrustes-alignment","display_name":"Procrustes alignment","score":0.8500000238418579},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.6259999871253967},{"id":"https://openalex.org/keywords/parameter-efficient-fine-tuning","display_name":"parameter-efficient fine-tuning","score":0.9049999713897705},{"id":"https://openalex.org/keywords/aggregation-error","display_name":"aggregation error","score":0.7089999914169312},{"id":"https://openalex.org/keywords/convergence-analysis","display_name":"convergence analysis","score":0.7390000224113464},{"id":"https://openalex.org/keywords/communication-efficiency","display_name":"communication efficiency","score":0.7940000295639038}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.76419997215271},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.6308000087738037},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.5605999827384949},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.517799973487854},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.48910000920295715},{"id":"https://openalex.org/C42355184","wikidata":"https://www.wikidata.org/wiki/Q1361088","display_name":"Matrix decomposition","level":3,"score":0.4814000129699707},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.461899995803833},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.4499000012874603},{"id":"https://openalex.org/C17349429","wikidata":"https://www.wikidata.org/wiki/Q1049914","display_name":"Matrix multiplication","level":3,"score":0.44690001010894775},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.44530001282691956},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.42329999804496765},{"id":"https://openalex.org/C2992525071","wikidata":"https://www.wikidata.org/wiki/Q50818671","display_name":"Federated learning","level":2,"score":0.37529999017715454},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.3734999895095825},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.35749998688697815},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3562000095844269},{"id":"https://openalex.org/C90673727","wikidata":"https://www.wikidata.org/wiki/Q901718","display_name":"Product (mathematics)","level":2,"score":0.350600004196167},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.34060001373291016},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.32420000433921814},{"id":"https://openalex.org/C56372850","wikidata":"https://www.wikidata.org/wiki/Q1050404","display_name":"Sparse matrix","level":3,"score":0.304500013589859},{"id":"https://openalex.org/C2775941552","wikidata":"https://www.wikidata.org/wiki/Q25212305","display_name":"Isolation (microbiology)","level":2,"score":0.3000999987125397},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2906000018119812},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2842000126838684},{"id":"https://openalex.org/C49555168","wikidata":"https://www.wikidata.org/wiki/Q176583","display_name":"Stochastic matrix","level":3,"score":0.2793000042438507},{"id":"https://openalex.org/C117884012","wikidata":"https://www.wikidata.org/wiki/Q94489","display_name":"n-gram","level":3,"score":0.27230000495910645},{"id":"https://openalex.org/C77246614","wikidata":"https://www.wikidata.org/wiki/Q1409400","display_name":"Gramian matrix","level":3,"score":0.267300009727478},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2646999955177307},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.2513999938964844}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2602.17095","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2602.17095","pdf_url":"https://arxiv.org/pdf/2602.17095","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,"listed_in":[],"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"},{"id":"doi:10.48550/arxiv.2602.17095","doi":"https://doi.org/10.48550/arxiv.2602.17095","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.17095","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,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2602.17095","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2602.17095","pdf_url":"https://arxiv.org/pdf/2602.17095","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,"listed_in":[],"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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7130717034.pdf","grobid_xml":"https://content.openalex.org/works/W7130717034.grobid-xml"},"referenced_works_count":11,"referenced_works":["https://openalex.org/W2799054028","https://openalex.org/W2963748441","https://openalex.org/W2965373594","https://openalex.org/W4229005866","https://openalex.org/W4322718191","https://openalex.org/W4327810158","https://openalex.org/W4362515116","https://openalex.org/W4385848613","https://openalex.org/W4401306886","https://openalex.org/W4416060654","https://openalex.org/W7117783816"],"related_works":["https://openalex.org/W7214720422","https://openalex.org/W4417307946","https://openalex.org/W7166840892","https://openalex.org/W4406658468","https://openalex.org/W4417526941","https://openalex.org/W7166864209","https://openalex.org/W4414858665","https://openalex.org/W7140073338","https://openalex.org/W7134803904","https://openalex.org/W4414813458"],"abstract_inverted_index":{"Parameter-efficient":[0],"fine-tuning":[1,29,49,116,125,171,201],"techniques":[2],"such":[3],"as":[4],"low-rank":[5,43,65,77,88,110,122],"adaptation":[6],"(LoRA)":[7],"enable":[8],"large":[9],"language":[10],"models":[11],"(LLMs)":[12],"to":[13,15,81,99,225],"adapt":[14],"downstream":[16,214],"tasks":[17],"efficiently.":[18],"Federated":[19],"learning":[20],"(FL)":[21],"further":[22],"facilitates":[23],"this":[24],"process":[25],"by":[26,158,211,223],"enabling":[27],"collaborative":[28],"across":[30,198],"distributed":[31],"clients":[32],"without":[33],"sharing":[34],"private":[35],"data.":[36],"However,":[37],"the":[38,63,70,73,83,91,104,132,144,149,155,166,179,187,220],"use":[39],"of":[40,53,75,134,137,181],"two":[41,51,64,76],"separate":[42],"matrices":[44],"in":[45,191],"LoRA":[46],"for":[47,124,173],"federated":[48,109,115],"introduces":[50],"types":[52],"challenges.":[54],"First,":[55],"aggregation":[56,112,145],"error":[57,146],"can":[58,97,142,218],"arise":[59],"from":[60],"separately":[61],"aggregating":[62],"matrices.":[66,89],"Second,":[67],"even":[68],"if":[69],"server":[71],"aggregates":[72,127],"product":[74],"matrices,":[78],"it":[79,96],"needs":[80],"decompose":[82],"aggregated":[84],"matrix":[85,123,130,133,168],"back":[86],"into":[87],"Since":[90],"decomposition":[92,100,156],"is":[93],"not":[94],"unique,":[95],"lead":[98],"drift.":[101],"To":[102],"tackle":[103],"aforementioned":[105],"challenges,":[106],"we":[107],"propose":[108],"Gram-matrix":[111],"(FLoRG),":[113],"a":[114,120,160,192],"framework":[117],"which":[118,164],"employs":[119],"single":[121],"and":[126,147,183,217],"its":[128,138],"Gram":[129],"(i.e.,":[131],"inner":[135],"products":[136],"column":[139],"vectors).":[140],"FLoRG":[141,182,205],"eliminate":[143],"reduce":[148,219],"communication":[150,221],"overhead.":[151],"It":[152],"also":[153],"minimizes":[154],"drift":[157],"introducing":[159],"Procrustes":[161,188],"alignment":[162,189],"approach":[163],"aligns":[165],"decomposed":[167],"between":[169],"consecutive":[170],"rounds":[172],"consistent":[174],"updates.":[175],"We":[176],"theoretically":[177],"analyze":[178],"convergence":[180,194],"prove":[184],"that":[185,204],"adopting":[186],"results":[190,197],"tighter":[193],"bound.":[195],"Experimental":[196],"multiple":[199],"LLM":[200],"benchmarks":[202],"demonstrate":[203],"outperforms":[206],"five":[207],"state-of-the-art":[208],"baseline":[209],"schemes":[210],"providing":[212],"higher":[213],"task":[215],"accuracy":[216],"overhead":[222],"up":[224],"2041$\\times$.":[226]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-02-21T00:00:00"}