{"id":"https://openalex.org/W7166885242","doi":"https://doi.org/10.18653/v1/2026.findings-acl.1625","title":"Rejection-to-Acceptance Transition: Model Editing-Based Jailbreak Backdoor Injection Not Limited to Few Output Tokens","display_name":"Rejection-to-Acceptance Transition: Model Editing-Based Jailbreak Backdoor Injection Not Limited to Few Output Tokens","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166885242","doi":"https://doi.org/10.18653/v1/2026.findings-acl.1625"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.1625","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1625","pdf_url":"https://aclanthology.org/2026.findings-acl.1625.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.1625.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139818002","display_name":"Shiji Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shiji Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139763315","display_name":"Min Cai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Min Cai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139750164","display_name":"Hao Xiong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hao Xiong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139833251","display_name":"Congyao Mei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Congyao Mei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088901640","display_name":"Haodong Zou","orcid":"https://orcid.org/0000-0003-3512-9683"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haodong Zou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139746240","display_name":"Shicheng Tan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shicheng Tan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139733677","display_name":"Jie Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jie Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139842170","display_name":"Fulan Qian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fulan Qian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139772457","display_name":"Shu Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shu Zhao","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":0.87363017,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"32463","last_page":"32477"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3122999966144562,"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":0.3122999966144562,"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/T11424","display_name":"Security and Verification in Computing","score":0.1281999945640564,"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/T10743","display_name":"Software Testing and Debugging Techniques","score":0.05849999934434891,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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/backdoor","display_name":"Backdoor","score":0.7084000110626221},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.3077999949455261},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.2567000091075897},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.2526000142097473}],"concepts":[{"id":"https://openalex.org/C2781045450","wikidata":"https://www.wikidata.org/wiki/Q254569","display_name":"Backdoor","level":2,"score":0.7084000110626221},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5432000160217285},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.46720001101493835},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.33379998803138733},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.325300008058548},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.3077999949455261},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.29440000653266907},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.2635999917984009},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2567000091075897},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2526000142097473},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2517000138759613}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.1625","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1625","pdf_url":"https://aclanthology.org/2026.findings-acl.1625.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.1625","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1625","pdf_url":"https://aclanthology.org/2026.findings-acl.1625.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.511553943157196}],"awards":[{"id":"https://openalex.org/G7510905396","display_name":null,"funder_award_id":"GXXT-2023-050","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7654974264","display_name":null,"funder_award_id":"62476003","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G8365206623","display_name":null,"funder_award_id":"2408055US006","funder_id":"https://openalex.org/F4320334897","funder_display_name":"Natural Science Foundation of Anhui Province"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320321435","display_name":"Anhui University","ror":"https://ror.org/05th6yx34"},{"id":"https://openalex.org/F4320334897","display_name":"Natural Science Foundation of Anhui Province","ror":null}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166885242.pdf","grobid_xml":"https://content.openalex.org/works/W7166885242.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Model":[0],"editing-based":[1],"jailbreak":[2,67,95,116,166],"backdoor":[3,68,100],"attacks":[4,69],"against":[5],"LLMs":[6,147,170],"have":[7],"gained":[8],"attention":[9],"for":[10,134],"being":[11],"lightweight,":[12],"enabling":[13],"vulnerability":[14,150],"discovery":[15],"in":[16],"LLMs.Existing":[17],"methods":[18],"are":[19],"implemented":[20],"by":[21,70],"binding":[22,80],"backdoors":[23,96],"to":[24,37,81,93,107,112,151,176],"predefined":[25],"phrases":[26],"as":[27,56],"first":[28],"few":[29,83],"output":[30,84],"tokens,":[31],"inducing":[32],"the":[33,47,102,143,178],"LLM's":[34],"next-token":[35],"prediction":[36],"produce":[38],"continuous":[39],"responses.However,":[40],"their":[41],"effectiveness":[42],"is":[43],"heavily":[44],"dependent":[45],"on":[46],"number":[48,58],"of":[49,181],"bound":[50],"phrases,":[51],"with":[52],"attack":[53],"costs":[54],"rising":[55],"this":[57,60],"increases.In":[59],"work,":[61],"we":[62,86,127],"propose":[63,87],"JEST,":[64],"which":[65],"achieves":[66],"hijacking":[71],"LLM":[72,103],"representations":[73],"into":[74,97],"a":[75,82,88,130],"acceptance":[76,108,123],"domain":[77,106,131],"rather":[78],"than":[79],"tokens.Specifically,":[85],"representation":[89,144],"transitionguided":[90],"model":[91,161],"editing":[92,162],"inject":[94],"LLMs.The":[98],"activated":[99],"transitions":[101],"from":[104],"rejection":[105,121],"domain,":[109],"causing":[110],"it":[111],"accept":[113],"and":[114,122,171],"generate":[115],"behavior.To":[117],"clearly":[118],"distinguish":[119],"between":[120],"domains":[124,141],"within":[125,142],"LLMs,":[126],"also":[128,173],"design":[129],"modeling":[132],"strategy":[133],"JEST":[135,158],"that":[136,157],"models":[137],"these":[138],"two":[139],"opposing":[140],"space.Additionally,":[145],"JEST-hijacked":[146],"exhibit":[148],"greater":[149],"direct":[152],"prompt":[153],"attacks.Experimental":[154],"results":[155],"show":[156],"outperforms":[159],"existing":[160],"methods,":[163],"demonstrating":[164],"stronger":[165],"capabilities":[167],"across":[168],"various":[169],"datasets.We":[172],"provide":[174],"analysis":[175],"explore":[177],"safety":[179],"boundary":[180],"LLM.":[182]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}