{"id":"https://openalex.org/W7101440805","doi":"https://doi.org/10.1109/tifs.2025.3626228","title":"Disguise While Defense: Avoid Refusal Responses in LLM\u2019s Defense via a Multi-Agent Attacker\u2013Disguiser Game","display_name":"Disguise While Defense: Avoid Refusal Responses in LLM\u2019s Defense via a Multi-Agent Attacker\u2013Disguiser Game","publication_year":2025,"publication_date":"2025-10-28","ids":{"openalex":"https://openalex.org/W7101440805","doi":"https://doi.org/10.1109/tifs.2025.3626228"},"language":null,"primary_location":{"id":"doi:10.1109/tifs.2025.3626228","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tifs.2025.3626228","pdf_url":null,"source":{"id":"https://openalex.org/S61310614","display_name":"IEEE Transactions on Information Forensics and Security","issn_l":"1556-6013","issn":["1556-6013","1556-6021"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Information Forensics and Security","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Qianqiao Xu","orcid":"https://orcid.org/0009-0009-6501-8686"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qianqiao Xu","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":"https://orcid.org/0009-0009-6501-8686","affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhiliang Tian","orcid":"https://orcid.org/0000-0002-8906-5198"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhiliang Tian","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":"https://orcid.org/0000-0002-8906-5198","affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hongyan Wu","orcid":null},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongyan Wu","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhen Huang","orcid":"https://orcid.org/0000-0003-4819-373X"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhen Huang","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":"https://orcid.org/0000-0003-4819-373X","affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Dong Liang","orcid":"https://orcid.org/0000-0001-5673-6605"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dong Liang","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":"https://orcid.org/0000-0001-5673-6605","affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yiping Song","orcid":"https://orcid.org/0000-0002-3277-6351"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yiping Song","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":"https://orcid.org/0000-0002-3277-6351","affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Ziyi Pan","orcid":null},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ziyi Pan","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Feng Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Feng Liu","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"last","author":{"id":null,"display_name":"Dongsheng Li","orcid":"https://orcid.org/0000-0001-9743-2034"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dongsheng Li","raw_affiliation_strings":["National University of Defense Technology, Changsha, Hunan, China"],"raw_orcid":"https://orcid.org/0000-0001-9743-2034","affiliations":[{"raw_affiliation_string":"National University of Defense Technology, Changsha, Hunan, China","institution_ids":["https://openalex.org/I170215575"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I170215575"],"apc_list":null,"apc_paid":null,"fwci":11.1372,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.9819967,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":"21","issue":null,"first_page":"1006","last_page":"1020"},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T13712","display_name":"Kantian Philosophy and Modern Interpretations","score":0.11999999731779099,"subfield":{"id":"https://openalex.org/subfields/1211","display_name":"Philosophy"},"field":{"id":"https://openalex.org/fields/12","display_name":"Arts and Humanities"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T13712","display_name":"Kantian Philosophy and Modern Interpretations","score":0.11999999731779099,"subfield":{"id":"https://openalex.org/subfields/1211","display_name":"Philosophy"},"field":{"id":"https://openalex.org/fields/12","display_name":"Arts and Humanities"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T14059","display_name":"Nietzsche, Schopenhauer, and Hegel","score":0.08410000056028366,"subfield":{"id":"https://openalex.org/subfields/1211","display_name":"Philosophy"},"field":{"id":"https://openalex.org/fields/12","display_name":"Arts and Humanities"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T13660","display_name":"Philosophical and Historical Studies","score":0.032999999821186066,"subfield":{"id":"https://openalex.org/subfields/3312","display_name":"Sociology and Political Science"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5993000268936157},{"id":"https://openalex.org/keywords/nash-equilibrium","display_name":"Nash equilibrium","score":0.5666000247001648},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.5634999871253967},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.42500001192092896},{"id":"https://openalex.org/keywords/complete-information","display_name":"Complete information","score":0.4244999885559082},{"id":"https://openalex.org/keywords/adversary","display_name":"Adversary","score":0.3797999918460846},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.33980000019073486},{"id":"https://openalex.org/keywords/game-theory","display_name":"Game theory","score":0.3278999924659729}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7824000120162964},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.7231000065803528},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5993000268936157},{"id":"https://openalex.org/C46814582","wikidata":"https://www.wikidata.org/wiki/Q23389","display_name":"Nash equilibrium","level":2,"score":0.5666000247001648},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.5634999871253967},{"id":"https://openalex.org/C108827166","wikidata":"https://www.wikidata.org/wiki/Q175975","display_name":"Internet privacy","level":1,"score":0.5530999898910522},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.42500001192092896},{"id":"https://openalex.org/C113336015","wikidata":"https://www.wikidata.org/wiki/Q574010","display_name":"Complete information","level":2,"score":0.4244999885559082},{"id":"https://openalex.org/C41065033","wikidata":"https://www.wikidata.org/wiki/Q2825412","display_name":"Adversary","level":2,"score":0.3797999918460846},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.33980000019073486},{"id":"https://openalex.org/C177142836","wikidata":"https://www.wikidata.org/wiki/Q44455","display_name":"Game theory","level":2,"score":0.3278999924659729},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3174000084400177},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.3154999911785126},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3131999969482422},{"id":"https://openalex.org/C149728462","wikidata":"https://www.wikidata.org/wiki/Q751319","display_name":"Minimax","level":2,"score":0.31299999356269836},{"id":"https://openalex.org/C32407928","wikidata":"https://www.wikidata.org/wiki/Q2733833","display_name":"Best response","level":3,"score":0.31189998984336853},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.3052999973297119},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.2761000096797943},{"id":"https://openalex.org/C21593369","wikidata":"https://www.wikidata.org/wiki/Q1032176","display_name":"Countermeasure","level":2,"score":0.2736000120639801},{"id":"https://openalex.org/C137822555","wikidata":"https://www.wikidata.org/wiki/Q2587068","display_name":"Information sensitivity","level":2,"score":0.27070000767707825},{"id":"https://openalex.org/C123650614","wikidata":"https://www.wikidata.org/wiki/Q282491","display_name":"Strategic dominance","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C2778403875","wikidata":"https://www.wikidata.org/wiki/Q20312394","display_name":"Adversarial machine learning","level":3,"score":0.2687999904155731},{"id":"https://openalex.org/C88959737","wikidata":"https://www.wikidata.org/wiki/Q1546627","display_name":"Strategy","level":3,"score":0.26820001006126404},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2554999887943268}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tifs.2025.3626228","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tifs.2025.3626228","pdf_url":null,"source":{"id":"https://openalex.org/S61310614","display_name":"IEEE Transactions on Information Forensics and Security","issn_l":"1556-6013","issn":["1556-6013","1556-6021"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Information Forensics and Security","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Reduced inequalities","score":0.45977848768234253,"id":"https://metadata.un.org/sdg/10"}],"awards":[{"id":"https://openalex.org/G2511849395","display_name":null,"funder_award_id":"62376284","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W1542941925","https://openalex.org/W2031098375","https://openalex.org/W2058111864","https://openalex.org/W2120327309","https://openalex.org/W2799899844","https://openalex.org/W3035207248","https://openalex.org/W3110979110","https://openalex.org/W3134354193","https://openalex.org/W3142849873","https://openalex.org/W4205923021","https://openalex.org/W4294214983","https://openalex.org/W4362578056","https://openalex.org/W4389519367","https://openalex.org/W4389617257","https://openalex.org/W4390742939","https://openalex.org/W4394744014","https://openalex.org/W4400315206","https://openalex.org/W4401042259","https://openalex.org/W4401996408","https://openalex.org/W4402217583","https://openalex.org/W4402671676","https://openalex.org/W4402671828","https://openalex.org/W4405181600","https://openalex.org/W4405464763","https://openalex.org/W4410609100","https://openalex.org/W4411113449","https://openalex.org/W4412944837","https://openalex.org/W4415797457"],"related_works":[],"abstract_inverted_index":{"With":[0],"the":[1,23,62,67,77,108,121,159,166,171,188,199,210,230,261,266,277,286,293],"enhanced":[2],"performance":[3],"of":[4,19,66,223],"large":[5],"language":[6,11],"models":[7],"(LLMs)":[8],"on":[9],"natural":[10],"processing":[12],"tasks,":[13],"potential":[14],"moral":[15],"and":[16,34,85,101,115,145,152,202,217,248,251,296],"ethical":[17],"issues":[18],"LLMs":[20,31,50,68],"arise":[21],"in":[22,280],"Web":[24],"service.":[25],"Malicious":[26],"attackers":[27,84],"exist":[28],"who":[29],"induce":[30],"to":[32,76,107,139,149,154,169,186,191,227,264,270],"jailbreak":[33,243],"generate":[35],"information":[36,40,87,119],"containing":[37],"illegal,":[38],"privacy-invasive":[39],"through":[41,69],"techniques":[42,56],"such":[43,57],"as":[44,58],"prompt":[45,193,218],"engineering.":[46],"As":[47],"a":[48],"result,":[49],"counter":[51],"malicious":[52,246],"attackers\u2019":[53],"attacks":[54],"using":[55],"safety":[59],"alignment.":[60],"However,":[61],"strong":[63],"defense":[64,96,267,299],"mechanism":[65],"rejection":[70],"replies,":[71],"which":[72,104,240],"are":[73],"typically":[74],"unrelated":[75],"attacker\u2019s":[78,294],"query,":[79],"is":[80,226],"easily":[81],"identified":[82],"by":[83],"leaks":[86],"that":[88,98,134,257],"strengthens":[89,260],"their":[90],"capabilities.":[91],"We":[92],"propose":[93],"an":[94,129,237],"LLM-driven":[95],"task":[97],"generates":[99],"safe":[100,250],"disguised":[102,252],"responses":[103],"remain":[105],"relevant":[106],"attack":[109,281],"request":[110],"while":[111,284],"concealing":[112],"defensive":[113],"intent":[114,268],"withholding":[116],"any":[117],"useful":[118],"from":[120,245],"attacker.":[122],"To":[123,157,177,197],"achieve":[124],"this":[125,224],"task,":[126],"we":[127,164,181,207,235],"construct":[128],"LLM-based":[130,137,238],"multi-agent":[131],"attacker-disguiser":[132],"game":[133,225],"trains":[135],"two":[136],"players":[138,172,211],"select":[140,192],"in-context":[141],"learning":[142],"(ICL)":[143],"examples":[144,201],"strategies,":[146],"guiding":[147],"them":[148],"better":[150],"generation":[151,161],"approach":[153],"Nash":[155,231],"equilibrium.":[156,232],"enhance":[158],"LLMs\u2019":[160,205],"via":[162],"ICL,":[163],"apply":[165],"Minimax":[167],"Q-learning":[168],"optimize":[170],"for":[173,195,204],"choosing":[174],"ICL":[175,200,213],"examples.":[176],"further":[178],"refine":[179],"generation,":[180,206],"use":[182],"policy":[183],"gradient":[184],"(PG)":[185],"train":[187,209],"strategy":[189,219],"selectors":[190],"strategies":[194],"LLMs.":[196],"improve":[198],"prompts":[203,244],"iteratively":[208],"with":[212],"sample":[214],"selection":[215],"training":[216],"training.":[220],"The":[221,254],"goal":[222],"arrive":[228],"at":[229],"Through":[233],"training,":[234],"obtain":[236],"defender":[239],"can":[241],"receive":[242],"users":[247],"output":[249],"responses.":[253],"experiments":[255],"reveal":[256],"our":[258],"method":[259,274],"LLM\u2019s":[262],"ability":[263],"disguise":[265],"compared":[269],"other":[271],"methods.":[272],"Our":[273],"also":[275],"minimizes":[276],"attacker-perceived":[278],"drop":[279],"success":[282],"rate,":[283],"delivering":[285],"largest":[287],"actual":[288],"reduction.":[289],"This":[290],"effectively":[291],"misleads":[292],"evaluation":[295],"enhances":[297],"practical":[298],"utility.":[300]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-28T00:00:00"}