{"id":"https://openalex.org/W7166862841","doi":"https://doi.org/10.18653/v1/2026.findings-acl.925","title":"RAMP: Risk-Aware Multi-Turn Planning for Jailbreak Red-Teaming","display_name":"RAMP: Risk-Aware Multi-Turn Planning for Jailbreak Red-Teaming","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166862841","doi":"https://doi.org/10.18653/v1/2026.findings-acl.925"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.925","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.925","pdf_url":"https://aclanthology.org/2026.findings-acl.925.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.925.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5026006304","display_name":"Yize Liu","orcid":"https://orcid.org/0009-0004-4775-996X"},"institutions":[{"id":"https://openalex.org/I4210142167","display_name":"DigitalSpace (United States)","ror":"https://ror.org/04h1th525","country_code":"US","type":"company","lineage":["https://openalex.org/I4210142167"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yize Liu","raw_affiliation_strings":["School of Computer and Cyberspace Security , Communication University of China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer and Cyberspace Security , Communication University of China","institution_ids":["https://openalex.org/I4210142167"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139740058","display_name":"Yunyun Hou","orcid":null},"institutions":[{"id":"https://openalex.org/I4210142167","display_name":"DigitalSpace (United States)","ror":"https://ror.org/04h1th525","country_code":"US","type":"company","lineage":["https://openalex.org/I4210142167"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yunyun Hou","raw_affiliation_strings":["School of Computer and Cyberspace Security , Communication University of China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer and Cyberspace Security , Communication University of China","institution_ids":["https://openalex.org/I4210142167"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5028395959","display_name":"Aina Sui","orcid":"https://orcid.org/0009-0009-5756-3983"},"institutions":[{"id":"https://openalex.org/I4210142167","display_name":"DigitalSpace (United States)","ror":"https://ror.org/04h1th525","country_code":"US","type":"company","lineage":["https://openalex.org/I4210142167"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Aina Sui","raw_affiliation_strings":["School of Computer and Cyberspace Security , Communication University of China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer and Cyberspace Security , Communication University of China","institution_ids":["https://openalex.org/I4210142167"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210142167"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.82038511,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"18553","last_page":"18564"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10906","display_name":"AI-based Problem Solving and Planning","score":0.5410000085830688,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.5410000085830688,"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/T11574","display_name":"Artificial Intelligence in Games","score":0.030899999663233757,"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/T10927","display_name":"Access Control and Trust","score":0.017899999395012856,"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/process","display_name":"Process (computing)","score":0.2906000018119812},{"id":"https://openalex.org/keywords/work","display_name":"Work (physics)","score":0.25850000977516174},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.24210000038146973},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.2298000007867813},{"id":"https://openalex.org/keywords/relation","display_name":"Relation (database)","score":0.22470000386238098}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3467999994754791},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2906000018119812},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.2768000066280365},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.25850000977516174},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.24210000038146973},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2298000007867813},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.226500004529953},{"id":"https://openalex.org/C25343380","wikidata":"https://www.wikidata.org/wiki/Q277521","display_name":"Relation (database)","level":2,"score":0.22470000386238098},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.22339999675750732},{"id":"https://openalex.org/C2776505523","wikidata":"https://www.wikidata.org/wiki/Q4785468","display_name":"Plan (archaeology)","level":2,"score":0.22310000658035278}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.925","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.925","pdf_url":"https://aclanthology.org/2026.findings-acl.925.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.925","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.925","pdf_url":"https://aclanthology.org/2026.findings-acl.925.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":[{"display_name":"Peace, Justice and strong institutions","score":0.4431549608707428,"id":"https://metadata.un.org/sdg/16"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166862841.pdf","grobid_xml":"https://content.openalex.org/works/W7166862841.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Multi-turn":[0],"jailbreaking":[1,53],"is":[2],"a":[3,27,55,76,87,89,92,119],"critical":[4],"approach":[5],"for":[6],"evaluating":[7],"the":[8],"safety":[9,63],"of":[10],"large":[11],"language":[12],"models":[13],"(LLMs).However,":[14],"existing":[15],"methods":[16],"largely":[17],"rely":[18],"on":[19],"heuristic":[20],"strategies":[21,70],"or":[22],"trained":[23],"attack":[24,42,101],"agents,":[25],"lacking":[26],"unified":[28],"state-action":[29],"formulation":[30],"and":[31,37,74,82,91,106,117,126,136],"systematic":[32],"search":[33],"over":[34],"strategy":[35],"compositions,":[36],"often":[38],"struggling":[39],"to":[40],"balance":[41],"success":[43],"rate":[44],"with":[45],"query":[46],"overhead.We":[47],"propose":[48],"RAMP,":[49],"which":[50],"formulates":[51],"multi-turn":[52],"as":[54,71],"risk-aware":[56],"PDDL":[57],"planning":[58],"problem.Specifically,":[59],"we":[60],"characterize":[61],"dialogue":[62],"using":[64],"predicate-based":[65],"states,":[66],"abstract":[67],"common":[68],"jailbreak":[69],"high-level":[72],"actions,":[73],"introduce":[75],"closed-loop":[77],"framework":[78],"that":[79,96,131],"iteratively":[80],"plans":[81],"executes":[83],"each":[84],"turn":[85,115],"via":[86],"Judge,":[88],"Transitioner,":[90],"Planner.Experimental":[93],"results":[94],"show":[95,130],"RAMP":[97],"achieves":[98],"consistently":[99],"strong":[100],"performance":[102],"across":[103,139],"both":[104],"open-source":[105],"closed-source":[107],"target":[108],"models,":[109],"while":[110],"remaining":[111],"effective":[112],"under":[113],"stricter":[114],"budgets":[116],"yielding":[118],"favorable":[120],"efficiencyeffectiveness":[121],"trade-off.Ablation":[122],"studies,":[123],"interpretability":[124],"analyses,":[125],"extended":[127],"experiments":[128],"further":[129],"multi-step":[132],"planning,":[133],"clue":[134],"accumulation,":[135],"consistent":[137],"findings":[138],"evaluator":[140],"settings":[141],"are":[142],"key":[143],"factors":[144],"underlying":[145],"these":[146],"gains.":[147]},"counts_by_year":[],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2026-07-02T00:00:00"}