{"id":"https://openalex.org/W7166847087","doi":"https://doi.org/10.18653/v1/2026.acl-long.357","title":"Enhancing the Transferability of Jailbreak Attacks on Large Language Models via Exploiting Reparameterization Invariance","display_name":"Enhancing the Transferability of Jailbreak Attacks on Large Language Models via Exploiting Reparameterization Invariance","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166847087","doi":"https://doi.org/10.18653/v1/2026.acl-long.357"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.357","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.357","pdf_url":"https://aclanthology.org/2026.acl-long.357.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.357.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139799187","display_name":"Ao Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ao Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034521535","display_name":"Xinghao Yang","orcid":"https://orcid.org/0000-0001-6487-3183"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xinghao Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040047825","display_name":"Yongshun Gong","orcid":"https://orcid.org/0000-0003-3948-4471"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yongshun Gong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139768460","display_name":"Wei Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5091703383","display_name":"Baodi Liu","orcid":"https://orcid.org/0000-0002-1408-5514"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bao-di Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139772462","display_name":"Weifeng Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Weifeng Liu","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.83689754,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"7854","last_page":"7865"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.20630000531673431,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.20630000531673431,"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/T12380","display_name":"Authorship Attribution and Profiling","score":0.18639999628067017,"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/T12262","display_name":"Hate Speech and Cyberbullying Detection","score":0.12960000336170197,"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/transferability","display_name":"Transferability","score":0.3723999857902527},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.31940001249313354},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.290800005197525},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.2671999931335449},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.2574000060558319}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6414999961853027},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.42809998989105225},{"id":"https://openalex.org/C61272859","wikidata":"https://www.wikidata.org/wiki/Q7834031","display_name":"Transferability","level":3,"score":0.3723999857902527},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3571000099182129},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3361000120639801},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3330000042915344},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.31940001249313354},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.290800005197525},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.2671999931335449},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.26499998569488525},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2637999951839447},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2574000060558319},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.25}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.357","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.357","pdf_url":"https://aclanthology.org/2026.acl-long.357.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.357","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.357","pdf_url":"https://aclanthology.org/2026.acl-long.357.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10","score":0.41310274600982666}],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166847087.pdf","grobid_xml":"https://content.openalex.org/works/W7166847087.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Jailbreak":[0,51],"attacks":[1,17,30],"serve":[2],"as":[3],"a":[4,53],"pivotal":[5],"technique":[6],"for":[7],"evaluating":[8],"the":[9,124],"safety":[10],"alignment":[11],"of":[12,98],"Large":[13],"language":[14,89],"models.Current":[15],"token-level":[16,64],"have":[18],"shown":[19],"remarkable":[20],"efficacy":[21],"on":[22,40],"open-source":[23],"models":[24,90],"by":[25,71,133],"leveraging":[26],"gradient-based":[27],"optimization.However,":[28],"these":[29],"suffer":[31],"from":[32],"poor":[33],"cross-model":[34,61,125],"transferability,":[35],"severely":[36],"limiting":[37],"their":[38,101],"utility":[39],"proprietary":[41],"ones.To":[42],"address":[43],"this":[44],"limitation,":[45],"we":[46],"propose":[47],"Reparameterization":[48],"Invariance":[49],"Gradient-based":[50],"(RIGJ),":[52],"natural":[54,99],"gradient":[55],"based":[56],"framework":[57],"designed":[58],"to":[59,80,93],"improve":[60],"transferability.Unlike":[62],"prior":[63],"methods":[65],"whose":[66],"optimization":[67,112],"paths":[68],"are":[69,91],"constrained":[70],"model-specific":[72],"Euclidean":[73],"geometry,":[74],"RIGJ":[75],"defines":[76],"update":[77],"directions":[78],"according":[79],"differences":[81],"in":[82,141],"output":[83,102],"distributions":[84,103],"rather":[85],"than":[86],"parameter-space":[87],"distances.Since":[88],"trained":[92],"capture":[94],"similar":[95],"dependency":[96],"structures":[97],"language,":[100],"share":[104],"common":[105],"geometry":[106],"across":[107],"architectures,":[108],"yielding":[109],"intrinsically":[110],"model-agnostic":[111],"trajectories":[113],"and":[114,129,135],"substantially":[115],"stronger":[116],"jailbreak":[117],"transferability.Extensive":[118],"experiments":[119],"demonstrate":[120],"superior":[121],"performance,":[122],"increasing":[123],"Attack":[126],"Success":[127],"Rate":[128],"Average":[130],"Harmfulness":[131],"Score":[132],"14.9%":[134],"1.23,":[136],"respectively.Our":[137],"code":[138],"is":[139],"provided":[140],"https://github.com/nohuma/AISafety_":[142],"transfer_jailbreak_RIGJ_2026.":[143]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}