{"id":"https://openalex.org/W7213402201","doi":"https://doi.org/10.48550/arxiv.2609.16977","title":"Structural Negative Transfer in Federated Graph Neural Networks: Diagnosis, Causal Investigation, and the Limits of Divergence-Aware Mitigation","display_name":"Structural Negative Transfer in Federated Graph Neural Networks: Diagnosis, Causal Investigation, and the Limits of Divergence-Aware Mitigation","publication_year":2026,"publication_date":"2026-09-15","ids":{"openalex":"https://openalex.org/W7213402201","doi":"https://doi.org/10.48550/arxiv.2609.16977"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.16977","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16977","pdf_url":"https://arxiv.org/pdf/2609.16977","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/2609.16977","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5126327770","display_name":"Chethana Prasad Kabgere","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kabgere, Chethana Prasad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5108721678","display_name":"S. S. Shylaja","orcid":"https://orcid.org/0000-0003-2628-8973","observed_orcids":["https://orcid.org/0000-0003-2628-8973"]},"institutions":[{"id":"https://openalex.org/I196608512","display_name":"PES University","ror":"https://ror.org/05m169e78","country_code":"IN","type":"education","lineage":["https://openalex.org/I196608512"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"SS, Shylaja","raw_affiliation_strings":["PES University , Bengaluru , India"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"PES University , Bengaluru , India","institution_ids":["https://openalex.org/I196608512"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"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":0.9994999766349792,"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":0.9994999766349792,"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/T11273","display_name":"Advanced Graph Neural Networks","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"}},{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.0,"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-graph-neural-networks","display_name":"federated graph neural networks","score":0.9150000214576721},{"id":"https://openalex.org/keywords/structural-divergence","display_name":"structural divergence","score":0.6850000023841858},{"id":"https://openalex.org/keywords/citation-networks","display_name":"citation networks","score":0.7070000171661377},{"id":"https://openalex.org/keywords/causal-inference","display_name":"causal inference","score":0.5379999876022339},{"id":"https://openalex.org/keywords/structural-statistics","display_name":"structural statistics","score":0.3610000014305115}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6686999797821045},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.5544999837875366},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.47690001130104065},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.4203000068664551},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.40470001101493835},{"id":"https://openalex.org/C207390915","wikidata":"https://www.wikidata.org/wiki/Q1230525","display_name":"Divergence (linguistics)","level":2,"score":0.3986999988555908},{"id":"https://openalex.org/C190470478","wikidata":"https://www.wikidata.org/wiki/Q2370229","display_name":"Invariant (physics)","level":2,"score":0.38449999690055847},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.37439998984336853},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35600000619888306},{"id":"https://openalex.org/C132964779","wikidata":"https://www.wikidata.org/wiki/Q2110223","display_name":"Raw data","level":2,"score":0.35409998893737793},{"id":"https://openalex.org/C2777363581","wikidata":"https://www.wikidata.org/wiki/Q15098235","display_name":"Harm","level":2,"score":0.34540000557899475},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.3395000100135803},{"id":"https://openalex.org/C83665646","wikidata":"https://www.wikidata.org/wiki/Q42139305","display_name":"Feature vector","level":2,"score":0.33340001106262207},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3328999876976013},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3149000108242035},{"id":"https://openalex.org/C169345407","wikidata":"https://www.wikidata.org/wiki/Q8216221","display_name":"Uncorrelated","level":2,"score":0.3109000027179718},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.305400013923645},{"id":"https://openalex.org/C141513077","wikidata":"https://www.wikidata.org/wiki/Q378542","display_name":"Independent and identically distributed random variables","level":3,"score":0.29820001125335693},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.2825999855995178},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.27950000762939453},{"id":"https://openalex.org/C146380142","wikidata":"https://www.wikidata.org/wiki/Q1137726","display_name":"Directed graph","level":2,"score":0.27140000462532043},{"id":"https://openalex.org/C38754835","wikidata":"https://www.wikidata.org/wiki/Q2003238","display_name":"Strongly connected component","level":2,"score":0.26910001039505005},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.25589999556541443},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.25529998540878296}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.16977","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16977","pdf_url":"https://arxiv.org/pdf/2609.16977","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.2609.16977","doi":"https://doi.org/10.48550/arxiv.2609.16977","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.16977","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-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.16977","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.16977","pdf_url":"https://arxiv.org/pdf/2609.16977","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/W7213402201.pdf","grobid_xml":"https://content.openalex.org/works/W7213402201.grobid-xml"},"referenced_works_count":2,"referenced_works":["https://openalex.org/W2990789643","https://openalex.org/W4403344826"],"related_works":["https://openalex.org/W4385764191","https://openalex.org/W4382318655","https://openalex.org/W7202291912","https://openalex.org/W7117997324","https://openalex.org/W7211876100","https://openalex.org/W7166270991","https://openalex.org/W4405900653","https://openalex.org/W7130405636","https://openalex.org/W4407953592","https://openalex.org/W4399911909"],"abstract_inverted_index":{"Federated":[0,20],"learning":[1,46],"lets":[2],"multiple":[3],"participants":[4],"train":[5],"a":[6,28,65,107,117,179,224,229,244,254],"shared":[7,34,90],"model":[8,17],"without":[9],"pooling":[10],"raw":[11],"data,":[12],"by":[13,129,182],"exchanging":[14],"locally":[15],"trained":[16],"updates":[18],"instead.":[19],"averaging":[21,24],"assumes":[22],"that":[23,49,154,250],"local":[25],"models":[26],"is":[27,243,251],"reasonable":[29],"way":[30],"to":[31,69,92,150,178],"solve":[32],"one":[33],"problem":[35],"when":[36,215],"participants'":[37],"data":[38],"are":[39],"broadly":[40],"similar.":[41],"Work":[42],"on":[43],"non-IID":[44],"federated":[45],"has":[47],"shown":[48],"this":[50,147],"assumption":[51],"can":[52],"withstand":[53],"differences":[54],"in":[55,78,84],"label":[56,79],"and":[57,113,164],"feature":[58,81],"distributions.":[59],"We":[60,98],"ask":[61],"whether":[62],"it":[63],"survives":[64,242],"different":[66,96],"strain":[67],"specific":[68],"graph":[70],"neural":[71],"networks,":[72],"where":[73],"client":[74,120],"graphs":[75],"differ":[76],"not":[77,173,211,252],"or":[80],"distribution":[82],"but":[83,209],"structure":[85],"itself,":[86],"requiring":[87],"the":[88,100,183,186,213,238],"same":[89],"weights":[91],"operate":[93],"over":[94],"fundamentally":[95],"topologies.":[97],"call":[99],"resulting":[101],"harm":[102,214],"structural":[103,115,138],"negative":[104],"transfer.":[105],"In":[106,131],"federation":[108],"of":[109,167,185,219],"real":[110],"citation":[111],"networks":[112],"synthetic":[114],"proxies,":[116],"structurally":[118,231],"atypical":[119],"lost":[121],"more":[122,162],"than":[123],"half":[124],"its":[125],"achievable":[126],"accuracy":[127],"simply":[128],"joining.":[130],"an":[132],"initial":[133],"six-client":[134],"federation,":[135],"two":[136],"label-free":[137],"statistics":[139],"computable":[140],"before":[141],"training":[142],"were":[143],"strongly":[144],"associated":[145,158],"with":[146,159],"harm.":[148],"Expanding":[149],"twenty":[151],"clients":[152],"showed":[153],"degree":[155],"divergence":[156,171],"remained":[157],"harm,":[160],"although":[161],"weakly,":[163],"survived":[165],"removal":[166],"domain":[168],"contrast.":[169],"Spectral":[170],"did":[172,210],"replicate,":[174],"which":[175,237],"we":[176],"trace":[177],"confound":[180],"caused":[181],"composition":[184],"reference":[187],"pool":[188],"used":[189],"for":[190],"leave-one-out":[191],"statistics.":[192],"A":[193,202],"causal":[194],"intervention":[195],"isolating":[196],"topology":[197],"found":[198],"no":[199],"significant":[200],"effect.":[201],"degree-normalization":[203],"mechanism":[204],"held":[205],"across":[206],"twenty-four":[207],"seeds":[208],"explain":[212],"corrected.":[216],"The":[217],"best":[218],"five":[220],"candidate":[221],"fixes":[222],"beat":[223],"tuned":[225],"baseline":[226],"only":[227],"until":[228],"matched,":[230],"blind":[232],"control":[233],"was":[234],"applied,":[235],"after":[236],"gain":[239],"disappeared.":[240],"What":[241],"modest,":[245],"partially":[246],"replicated,":[247],"degree-specific":[248],"signal":[249],"yet":[253],"validated":[255],"predictor":[256],"at":[257],"scale.":[258]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-17T00:00:00"}