{"id":"https://openalex.org/W7224952170","doi":"https://doi.org/10.5281/zenodo.23801542","title":"Beyond Representation Similarity: Exchange-Diagram Residuals for Diagnosing Neural Network Module Compatibility: A Computational Proof-of-Concept Study","display_name":"Beyond Representation Similarity: Exchange-Diagram Residuals for Diagnosing Neural Network Module Compatibility: A Computational Proof-of-Concept Study","publication_year":2026,"publication_date":"2026-10-30","ids":{"openalex":"https://openalex.org/W7224952170","doi":"https://doi.org/10.5281/zenodo.23801542"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23801542","doi":"https://doi.org/10.5281/zenodo.23801542","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23801542","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"volume":null,"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.23801542","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5143709963","display_name":"Yancheng Guo","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Yancheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5145615924","display_name":"Hu Yuanfei","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Yuanfei","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/T10320","display_name":"Neural Networks and Applications","score":0.42410001158714294,"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/T10320","display_name":"Neural Networks and Applications","score":0.42410001158714294,"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":0.22949999570846558,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.20350000262260437,"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/representational-similarity","display_name":"representational similarity","score":0.8399999737739563},{"id":"https://openalex.org/keywords/mismatch-detection","display_name":"mismatch detection","score":0.7269999980926514},{"id":"https://openalex.org/keywords/principal-component-analysis","display_name":"principal component analysis","score":0.6700000166893005},{"id":"https://openalex.org/keywords/centered-kernel-alignment","display_name":"centered kernel alignment","score":0.5889999866485596},{"id":"https://openalex.org/keywords/multilayer-perceptron","display_name":"multilayer perceptron","score":0.7409999966621399},{"id":"https://openalex.org/keywords/symmetric-group-s3","display_name":"symmetric group S3","score":1.0},{"id":"https://openalex.org/keywords/continual-learning","display_name":"continual learning","score":0.7770000100135803},{"id":"https://openalex.org/keywords/model-maintenance","display_name":"model maintenance","score":0.8119999766349792},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"fine-tuning","score":0.6050000190734863},{"id":"https://openalex.org/keywords/pruning","display_name":"pruning","score":0.828000009059906}],"concepts":[{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.7457000017166138},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6270999908447266},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.5443000197410583},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.5403000116348267},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.5379999876022339},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5056999921798706},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5016000270843506},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.48660001158714294},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4449999928474426},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.4399000108242035},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.43299999833106995},{"id":"https://openalex.org/C60908668","wikidata":"https://www.wikidata.org/wiki/Q690207","display_name":"Perceptron","level":3,"score":0.4311999976634979},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.41290000081062317},{"id":"https://openalex.org/C153874254","wikidata":"https://www.wikidata.org/wiki/Q115542","display_name":"Canonical correlation","level":2,"score":0.41130000352859497},{"id":"https://openalex.org/C203519979","wikidata":"https://www.wikidata.org/wiki/Q865360","display_name":"Jaccard index","level":3,"score":0.3774999976158142},{"id":"https://openalex.org/C179717631","wikidata":"https://www.wikidata.org/wiki/Q2991667","display_name":"Multilayer perceptron","level":3,"score":0.3716000020503998},{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.3644999861717224},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.3522999882698059},{"id":"https://openalex.org/C162838799","wikidata":"https://www.wikidata.org/wiki/Q596077","display_name":"Counterexample","level":2,"score":0.3285999894142151},{"id":"https://openalex.org/C27438332","wikidata":"https://www.wikidata.org/wiki/Q2873","display_name":"Principal component analysis","level":2,"score":0.32659998536109924},{"id":"https://openalex.org/C2780092901","wikidata":"https://www.wikidata.org/wiki/Q3433612","display_name":"Correlation coefficient","level":2,"score":0.31709998846054077},{"id":"https://openalex.org/C32946077","wikidata":"https://www.wikidata.org/wiki/Q618079","display_name":"Network analysis","level":2,"score":0.30979999899864197},{"id":"https://openalex.org/C191172861","wikidata":"https://www.wikidata.org/wiki/Q7899321","display_name":"Upstream (networking)","level":2,"score":0.2849999964237213},{"id":"https://openalex.org/C117220453","wikidata":"https://www.wikidata.org/wiki/Q5172842","display_name":"Correlation","level":2,"score":0.28220000863075256},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.28209999203681946},{"id":"https://openalex.org/C158980903","wikidata":"https://www.wikidata.org/wiki/Q7899329","display_name":"Upstream and downstream (DNA)","level":3,"score":0.271699994802475},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.25369998812675476},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.2524999976158142},{"id":"https://openalex.org/C2781311116","wikidata":"https://www.wikidata.org/wiki/Q83306","display_name":"Group (periodic table)","level":2,"score":0.2506999969482422}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23801542","doi":"https://doi.org/10.5281/zenodo.23801542","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23801542","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.23801543","doi":"https://doi.org/10.5281/zenodo.23801543","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23801543","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":"doi:10.5281/zenodo.23801542","doi":"https://doi.org/10.5281/zenodo.23801542","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23801542","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W3188789802","https://openalex.org/W7133209506","https://openalex.org/W2942810103","https://openalex.org/W7155075871","https://openalex.org/W2801366793","https://openalex.org/W3210904560","https://openalex.org/W2547222832","https://openalex.org/W2941548735","https://openalex.org/W7125825898","https://openalex.org/W4225552216"],"abstract_inverted_index":{"The":[0,171,336],"email":[1],"address":[2],"of":[3,118,226,338],"this":[4],"preprint:":[5],"yanchengguo423@gmail.com;":[6],"yuanfeihu_duke@outlook.com":[7],"Neural":[8],"network":[9,64,84,286],"representations":[10],"are":[11],"commonly":[12],"compared":[13],"using":[14,132,153],"similarity":[15,39],"measures":[16],"such":[17,123],"as":[18,280],"centered":[19],"kernel":[20],"alignment":[21],"(CKA)":[22],"and":[23,110,141,147,166,181,308,322,332,351],"singular":[24],"vector":[25],"canonical":[26],"correlation":[27],"analysis":[28,137],"(SVCCA).":[29],"Although":[30],"these":[31,326],"methods":[32],"characterize":[33],"relationships":[34],"between":[35,305],"internal":[36],"representations,":[37],"representational":[38,148,306],"does":[40],"not":[41,266,293],"necessarily":[42],"establish":[43,295],"whether":[44,74,96],"a":[45,63,97,127,133,296,303],"previously":[46],"trained":[47,156],"downstream":[48,99,245,255,309],"module":[49,85,100,287],"remains":[50],"compatible":[51],"with":[52,122,144,244,253],"an":[53,106],"updated":[54],"upstream":[55,107],"module.":[56],"This":[57,71],"distinction":[58,304],"becomes":[59],"particularly":[60],"important":[61],"when":[62],"undergoes":[65],"fine-tuning,":[66],"pruning,":[67],"or":[68],"function-preserving":[69],"transformations.":[70],"study":[72],"investigates":[73],"exchange-diagram":[75,129,276],"residuals":[76,241,250,277],"can":[77,101,242,251],"provide":[78],"structural":[79,282],"information":[80],"relevant":[81],"to":[82,208,221,231,294,301],"neural":[83,285],"compatibility.":[86,288,311],"We":[87,125],"distinguish":[88],"two":[89],"tasks:":[90],"(1)":[91],"mismatch":[92],"detection,":[93],"which":[94,114],"determines":[95],"frozen":[98],"still":[102],"be":[103],"reused":[104],"after":[105],"representation":[108],"update,":[109],"(2)":[111],"downstream-damage":[112,213],"prediction,":[113,214],"estimates":[115],"the":[116,162,167,200,224,262,345,348,353],"magnitude":[117],"performance":[119],"degradation":[120],"associated":[121],"reuse.":[124],"introduce":[126],"fixed-coordinate":[128,201],"residual,":[130],"calculated":[131],"reference":[134],"principal":[135,290],"component":[136,318],"(PCA)":[138],"coordinate":[139],"system,":[140],"compare":[142],"it":[143],"conventional":[145],"functional":[146,310],"diagnostics.":[149],"Experiments":[150],"were":[151],"conducted":[152],"multilayer":[154],"perceptrons":[155],"on":[157],"synthetic":[158],"group-operation":[159],"tasks":[160],"involving":[161],"cyclic":[163],"group":[164,169],"Z7":[165],"symmetric":[168],"S3.":[170],"confirmatory":[172],"experiment":[173,349],"included":[174],"40":[175],"random":[176],"seeds,":[177],"three":[178],"macro-representation":[179],"dimensions,":[180],"11":[182],"update":[183],"conditions,":[184],"producing":[185],"2,640":[186],"experimental":[187],"records.":[188],"Of":[189],"these,":[190],"1,920":[191],"satisfied":[192],"predefined":[193],"external-behavior-preservation":[194],"criteria.":[195],"In":[196],"grouped":[197],"cross-validation,":[198],"adding":[199],"residual":[202],"increased":[203,228],"mismatch-detection":[204],"AUROC":[205],"from":[206,219,229],"0.927":[207],"approximately":[209],"1.000.":[210],"For":[211],"continuous":[212],"mean":[215],"absolute":[216],"error":[217],"decreased":[218],"1.793":[220],"1.228,":[222],"while":[223,247],"coefficient":[225],"determination":[227],"0.784":[230],"0.885.":[232],"Two":[233],"deliberately":[234],"constructed":[235],"counterexamples":[236],"demonstrated":[237],"that":[238,261,275],"high":[239],"exchange":[240],"coexist":[243,252],"compatibility,":[246],"nearly":[248],"zero":[249],"severe":[254],"incompatibility.":[256],"Furthermore,":[257],"leave-one-mechanism-out":[258],"analyses":[259],"revealed":[260],"predictive":[263],"relationship":[264],"was":[265,341],"uniformly":[267],"transferable":[268],"across":[269],"intervention":[270],"mechanisms.":[271],"These":[272],"findings":[273],"suggest":[274],"may":[278],"serve":[279],"complementary":[281],"diagnostics":[283],"for":[284,343],"Their":[289],"contribution":[291],"is":[292],"universal":[297],"compatibility":[298],"criterion,":[299],"but":[300],"highlight":[302],"consistency":[307],"Potential":[312],"applications":[313,327],"include":[314],"modular":[315],"model":[316,323],"updating,":[317],"reuse,":[319],"continual":[320],"learning,":[321],"maintenance,":[324],"although":[325],"require":[328],"validation":[329],"in":[330],"larger":[331],"more":[333],"realistic":[334],"architectures.":[335],"use":[337],"AI:":[339],"AI":[340],"used":[342],"searching":[344],"resource,":[346],"writting":[347],"codes,":[350],"polishing":[352],"preprint.":[354]},"counts_by_year":[],"updated_date":"2026-11-01T11:40:01.658920","created_date":"2026-10-31T00:00:00"}