{"id":"https://openalex.org/W7131126583","doi":"https://doi.org/10.48550/arxiv.2602.17711","title":"Interpreting Multi-Branch Anti-Spoofing Architectures: Correlating Internal Strategy with Empirical Performance","display_name":"Interpreting Multi-Branch Anti-Spoofing Architectures: Correlating Internal Strategy with Empirical Performance","publication_year":2026,"publication_date":"2026-02-14","ids":{"openalex":"https://openalex.org/W7131126583","doi":"https://doi.org/10.48550/arxiv.2602.17711"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2602.17711","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.17711","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2602.17711","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5119981664","display_name":"Ivan Viakhirev","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Viakhirev, Ivan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114316560","display_name":"Kirill Borodin","orcid":"https://orcid.org/0009-0001-8203-1059"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Borodin, Kirill","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063243000","display_name":"Mikhail Gorodnichev","orcid":"https://orcid.org/0000-0003-1739-9831"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gorodnichev, Mikhail","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5084139711","display_name":"\u0413 \u041c \u041c\u043a\u0440\u0442\u0447\u044f\u043d","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mkrtchian, Grach","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5119981664"],"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/T11309","display_name":"Music and Audio Processing","score":0.25619998574256897,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T11309","display_name":"Music and Audio Processing","score":0.25619998574256897,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.10830000042915344,"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/T11241","display_name":"Advanced Malware Detection Techniques","score":0.09860000014305115,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/interpretability","display_name":"Interpretability","score":0.8916000127792358},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.529699981212616},{"id":"https://openalex.org/keywords/focus","display_name":"Focus (optics)","score":0.4081999957561493},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3741999864578247},{"id":"https://openalex.org/keywords/mode","display_name":"Mode (computer interface)","score":0.3718999922275543},{"id":"https://openalex.org/keywords/semantics","display_name":"Semantics (computer science)","score":0.3481000065803528},{"id":"https://openalex.org/keywords/spoofing-attack","display_name":"Spoofing attack","score":0.3149999976158142},{"id":"https://openalex.org/keywords/covariance","display_name":"Covariance","score":0.3133000135421753}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.8916000127792358},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7350000143051147},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5378000140190125},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.529699981212616},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.42579999566078186},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.4081999957561493},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3741999864578247},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.3718999922275543},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3481000065803528},{"id":"https://openalex.org/C167900197","wikidata":"https://www.wikidata.org/wiki/Q11081100","display_name":"Spoofing attack","level":2,"score":0.3149999976158142},{"id":"https://openalex.org/C178650346","wikidata":"https://www.wikidata.org/wiki/Q201984","display_name":"Covariance","level":2,"score":0.3133000135421753},{"id":"https://openalex.org/C28427503","wikidata":"https://www.wikidata.org/wiki/Q13580300","display_name":"Internal model","level":3,"score":0.3052999973297119},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.28700000047683716},{"id":"https://openalex.org/C2778565505","wikidata":"https://www.wikidata.org/wiki/Q2207566","display_name":"Spec#","level":2,"score":0.28540000319480896},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.2831999957561493},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.26269999146461487},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.2599000036716461},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.25609999895095825},{"id":"https://openalex.org/C166052673","wikidata":"https://www.wikidata.org/wiki/Q83021","display_name":"Empirical evidence","level":2,"score":0.2556999921798706},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.2540999948978424}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2602.17711","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.17711","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2602.17711","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.17711","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","score":0.7481865286827087,"id":"https://metadata.un.org/sdg/16"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Multi-branch":[0],"deep":[1],"neural":[2],"networks":[3],"like":[4],"AASIST3":[5,59],"achieve":[6],"state-of-the-art":[7],"comparable":[8],"performance":[9,168,197],"in":[10,161],"audio":[11],"anti-spoofing,":[12],"yet":[13],"their":[14],"internal":[15,184],"decision":[16],"dynamics":[17],"remain":[18],"opaque":[19],"compared":[20],"to":[21,91,107,139,166,188],"traditional":[22],"input-level":[23],"saliency":[24],"methods.":[25],"While":[26],"existing":[27],"interpretability":[28],"efforts":[29],"largely":[30],"focus":[31],"on":[32],"visualizing":[33],"input":[34],"artifacts,":[35],"the":[36,61,109,119,156],"way":[37],"individual":[38],"architectural":[39,185],"branches":[40,68],"cooperate":[41],"or":[42],"compete":[43],"under":[44],"different":[45],"spoofing":[46,116],"attacks":[47,117,171],"is":[48],"not":[49],"well":[50],"characterized.":[51],"This":[52],"paper":[53],"develops":[54],"a":[55,88,151],"framework":[56],"for":[57,170],"interpreting":[58],"at":[60],"component":[62],"level.":[63],"Intermediate":[64],"activations":[65],"from":[66,118,128],"fourteen":[67],"and":[69,103,173,177],"global":[70],"attention":[71],"modules":[72],"are":[73],"modeled":[74],"with":[75],"covariance":[76],"operators":[77],"whose":[78],"leading":[79,165],"eigenvalues":[80],"form":[81],"low-dimensional":[82],"spectral":[83],"signatures.":[84],"These":[85,180],"signatures":[86],"train":[87],"CatBoost":[89],"meta-classifier":[90],"generate":[92],"TreeSHAP-based":[93],"branch":[94],"attributions,":[95],"which":[96],"we":[97,123],"convert":[98],"into":[99],"normalized":[100],"contribution":[101],"shares":[102],"confidence":[104,160],"scores":[105],"(Cb)":[106],"quantify":[108],"model's":[110],"operational":[111,126],"strategy.":[112],"By":[113],"analyzing":[114],"13":[115],"ASVspoof":[120],"2019":[121],"benchmark,":[122],"identify":[124],"four":[125],"archetypes-ranging":[127],"Effective":[129],"Specialization":[130,153],"(e.g.,":[131,142],"A09,":[132],"Equal":[133],"Error":[134],"Rate":[135],"(EER)":[136],"0.04%,":[137],"C=1.56)":[138],"Ineffective":[140],"Consensus":[141],"A08,":[143],"EER":[144],"3.14%,":[145],"C=0.33).":[146],"Crucially,":[147],"our":[148],"analysis":[149],"exposes":[150],"Flawed":[152],"mode":[154],"where":[155],"model":[157],"places":[158],"high":[159],"an":[162],"incorrect":[163],"branch,":[164],"severe":[167],"degradation":[169],"A17":[172],"A18":[174],"(EER":[175],"14.26%":[176],"28.63%,":[178],"respectively).":[179],"quantitative":[181],"findings":[182],"link":[183],"strategy":[186],"directly":[187],"empirical":[189],"reliability,":[190],"highlighting":[191],"specific":[192],"structural":[193],"dependencies":[194],"that":[195],"standard":[196],"metrics":[198],"overlook.":[199]},"counts_by_year":[],"updated_date":"2026-02-24T06:20:37.074948","created_date":"2026-02-24T00:00:00"}