{"id":"https://openalex.org/W7213423337","doi":"https://doi.org/10.48550/arxiv.2609.16099","title":"SWB-DM: A Calibrated Sliced-Wasserstein-Barycenter Aggregator with Delayed-Momentum Caching for Byzantine-Robust Federated Learning under Partial Participation","display_name":"SWB-DM: A Calibrated Sliced-Wasserstein-Barycenter Aggregator with Delayed-Momentum Caching for Byzantine-Robust Federated Learning under Partial Participation","publication_year":2026,"publication_date":"2026-09-14","ids":{"openalex":"https://openalex.org/W7213423337","doi":"https://doi.org/10.48550/arxiv.2609.16099"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.16099","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16099","pdf_url":"https://arxiv.org/pdf/2609.16099","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":null,"license_id":null,"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/2609.16099","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5151920053","display_name":"Saranraj S","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"S, Saranraj","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5154175330","display_name":"Saranya M S","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"S, Saranya M","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5151898797","display_name":"Alex David S","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"S, Alex David","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5159894578","display_name":"Ajay Kumar A","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"A, Ajay Kumar","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":true,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9976000189781189,"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":0.9976000189781189,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.0012000000569969416,"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":0.0003000000142492354,"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/robust-aggregation","display_name":"robust aggregation","score":0.8149999976158142},{"id":"https://openalex.org/keywords/wasserstein-barycenter","display_name":"Wasserstein barycenter","score":0.9079999923706055},{"id":"https://openalex.org/keywords/partial-client-participation","display_name":"partial client participation","score":0.671999990940094},{"id":"https://openalex.org/keywords/median-aggregation","display_name":"median aggregation","score":0.42500001192092896},{"id":"https://openalex.org/keywords/krum","display_name":"Krum","score":0.8650000095367432},{"id":"https://openalex.org/keywords/trimmed-mean","display_name":"trimmed mean","score":0.4779999852180481},{"id":"https://openalex.org/keywords/convergence-bounds","display_name":"convergence bounds","score":0.4000000059604645}],"concepts":[{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.6948000192642212},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6263999938964844},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6013000011444092},{"id":"https://openalex.org/C180505990","wikidata":"https://www.wikidata.org/wiki/Q498267","display_name":"News aggregator","level":2,"score":0.5485000014305115},{"id":"https://openalex.org/C2992525071","wikidata":"https://www.wikidata.org/wiki/Q50818671","display_name":"Federated learning","level":2,"score":0.515500009059906},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.4934999942779541},{"id":"https://openalex.org/C2908647359","wikidata":"https://www.wikidata.org/wiki/Q2625603","display_name":"Population","level":2,"score":0.4424999952316284},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.4214000105857849},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.3889999985694885},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.3781000077724457},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.3391999900341034},{"id":"https://openalex.org/C8505890","wikidata":"https://www.wikidata.org/wiki/Q605095","display_name":"Budget constraint","level":2,"score":0.33340001106262207},{"id":"https://openalex.org/C66882249","wikidata":"https://www.wikidata.org/wiki/Q169336","display_name":"Homogeneous","level":2,"score":0.3248000144958496},{"id":"https://openalex.org/C149629883","wikidata":"https://www.wikidata.org/wiki/Q660926","display_name":"Fraction (chemistry)","level":2,"score":0.3122999966144562},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.29179999232292175},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2840000092983246},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.2824999988079071},{"id":"https://openalex.org/C67226441","wikidata":"https://www.wikidata.org/wiki/Q1665389","display_name":"Robust statistics","level":3,"score":0.2766000032424927},{"id":"https://openalex.org/C88611116","wikidata":"https://www.wikidata.org/wiki/Q957004","display_name":"Trim","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C48406656","wikidata":"https://www.wikidata.org/wiki/Q534112","display_name":"Martingale (probability theory)","level":2,"score":0.26669999957084656},{"id":"https://openalex.org/C3020318244","wikidata":"https://www.wikidata.org/wiki/Q4812187","display_name":"Large sample","level":2,"score":0.2662000060081482},{"id":"https://openalex.org/C60718061","wikidata":"https://www.wikidata.org/wiki/Q1414747","display_name":"Momentum (technical analysis)","level":2,"score":0.263700008392334},{"id":"https://openalex.org/C190470478","wikidata":"https://www.wikidata.org/wiki/Q2370229","display_name":"Invariant (physics)","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C116714509","wikidata":"https://www.wikidata.org/wiki/Q32278","display_name":"Truncated mean","level":3,"score":0.2526000142097473},{"id":"https://openalex.org/C42747912","wikidata":"https://www.wikidata.org/wiki/Q1048447","display_name":"Multiplicative function","level":2,"score":0.25200000405311584},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.251800000667572}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.16099","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16099","pdf_url":"https://arxiv.org/pdf/2609.16099","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.16099","doi":"https://doi.org/10.48550/arxiv.2609.16099","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.16099","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.16099","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16099","pdf_url":"https://arxiv.org/pdf/2609.16099","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":null,"license_id":null,"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/W7213423337.pdf","grobid_xml":"https://content.openalex.org/works/W7213423337.grobid-xml"},"referenced_works_count":1,"referenced_works":["https://openalex.org/W4417409718"],"related_works":["https://openalex.org/W7222062505","https://openalex.org/W4417409718","https://openalex.org/W7165896797","https://openalex.org/W7212066131","https://openalex.org/W6943895950","https://openalex.org/W4322771378","https://openalex.org/W7204205773","https://openalex.org/W7213785727","https://openalex.org/W7213595078","https://openalex.org/W3016256827"],"abstract_inverted_index":{"Robust":[0],"aggregation":[1],"methods":[2],"for":[3,264],"federated":[4],"learning":[5],"quietly":[6],"rest":[7],"on":[8],"a":[9,17,21,35,43,81,85,89,100,105,152,165,181,209,232,238],"fragile":[10],"assumption:":[11],"that":[12,50,58],"whoever":[13,135],"shows":[14,251],"up":[15],"in":[16],"given":[18],"round":[19,249],"is":[20,143],"fair":[22],"sample":[23,51],"of":[24,37,46,80],"the":[25,55,126,247],"full":[26,127],"population.":[27],"In":[28],"practice,":[29],"they":[30],"rarely":[31],"are.":[32],"When":[33],"only":[34],"handful":[36],"clients":[38],"participate":[39],"per":[40],"round,":[41,131],"even":[42],"modest":[44],"fraction":[45],"adversaries":[47],"can":[48],"dominate":[49],"and":[52,63,95,109,164,195],"silently":[53,186],"invalidate":[54],"finite-sample":[56],"guarantees":[57],"coordinate-wise":[59,177],"median,":[60],"Krum,":[61],"Bulyan,":[62],"trimmed":[64,90],"mean":[65],"all":[66,244],"depend":[67],"on.":[68],"We":[69],"introduce":[70],"SWB-DM":[71],"to":[72,115,137,180,205,246,268],"address":[73],"this":[74],"directly.":[75],"SWB":[76,221],"treats":[77],"each":[78,130],"slice":[79],"client":[82,128],"update":[83],"as":[84],"one-dimensional":[86],"distribution,":[87],"computes":[88],"Wasserstein":[91],"barycenter":[92],"across":[93,125],"clients,":[94],"recovers":[96],"coordinate":[97],"identity":[98],"via":[99],"medoid-based":[101],"gauge-fixing":[102],"step":[103],"--":[104,219,222,263],"heuristic":[106],"we":[107,169],"developed":[108],"do":[110],"not":[111,144],"claim":[112],"it":[113],"belongs":[114],"standard":[116],"optimal-transport":[117],"theory.":[118],"DeMoA-style":[119],"delayed":[120],"momentum":[121],"then":[122],"caches":[123],"updates":[124],"population":[129],"decoupling":[132],"robustness":[133],"from":[134],"happened":[136],"be":[138],"sampled.":[139],"Trim":[140],"ratio":[141],"calibration":[142],"cosmetic:":[145],"under-trimming":[146],"causes":[147],"collapse":[148],"at":[149],"corruption":[150],"levels":[151],"properly":[153],"calibrated":[154],"model":[155],"survives.":[156],"Across":[157],"448":[158],"CIFAR-10":[159,253],"configurations,":[160],"plus":[161],"CIFAR-100,":[162,259],"FEMNIST,":[163],"500-client":[166],"scalability":[167],"run,":[168],"find":[170],"several":[171],"mechanistically":[172],"distinct":[173],"failure":[174],"modes.":[175],"Even-sample":[176],"median":[178],"degrades":[179],"deterministic":[182],"wrong":[183],"answer.":[184],"Krum":[185],"violates":[187],"its":[188,252],"own":[189],"n":[190,200],"greater":[191,201],"than":[192,202,225],"2f+2":[193],"precondition":[194],"diverges":[196],"without":[197],"warning.":[198],"Bulyan's":[199],"or":[203],"equal":[204],"4f+3":[206],"threshold":[207],"produces":[208],"sharp":[210],"pass/fail":[211],"boundary.":[212],"On":[213,258],"attacks,":[214],"IPM":[215],"defeats":[216],"order-statistic":[217],"defenses":[218],"including":[220],"more":[223,262],"reliably":[224],"ALIE,":[226],"confirmed":[227],"through":[228],"delta-space":[229],"measurements":[230],"against":[231],"convergence":[233],"bound.":[234],"SWB-DM's":[235],"cache":[236],"carries":[237],"real":[239],"warm-up":[240],"cost,":[241],"but":[242],"extending":[243],"baselines":[245],"same":[248],"budget":[250],"gains":[254],"are":[255],"disproportionately":[256],"large.":[257],"FLTrust":[260],"benefits":[261],"reasons":[265],"entirely":[266],"unrelated":[267],"caching.":[269]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-17T00:00:00"}