{"id":"https://openalex.org/W4403051664","doi":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","title":"Crescendo Vulnerability in GPT o1 (Strawberry) Series: Bypassing Slow-thinking with the Single Turn Crescendo Attack","display_name":"Crescendo Vulnerability in GPT o1 (Strawberry) Series: Bypassing Slow-thinking with the Single Turn Crescendo Attack","publication_year":2024,"publication_date":"2024-10-02","ids":{"openalex":"https://openalex.org/W4403051664","doi":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1"},"language":"en","primary_location":{"id":"doi:10.36227/techrxiv.172788963.37426055/v1","doi":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","is_oa":true,"landing_page_url":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","pdf_url":"https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.172788963.37426055","source":null,"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"posted-content"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.172788963.37426055","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5107693708","display_name":"Alan Aqrawi","orcid":"https://orcid.org/0009-0002-2117-3894","observed_orcids":["https://orcid.org/0009-0002-2117-3894"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Alan Aqrawi","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0009-0002-2117-3894","affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5113413382","display_name":"Arian Abbasi","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Arian Abbasi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5107693708"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9994999766349792,"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":0.9994999766349792,"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/T10028","display_name":"Topic Modeling","score":9.999999747378752e-05,"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":9.999999747378752e-05,"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/large-language-model-security","display_name":"large language model security","score":0.593999981880188},{"id":"https://openalex.org/keywords/adversarial-prompts","display_name":"adversarial prompts","score":0.6380000114440918},{"id":"https://openalex.org/keywords/slow-thinking","display_name":"slow thinking","score":0.46799999475479126},{"id":"https://openalex.org/keywords/familiarity-effects","display_name":"familiarity effects","score":0.7960000038146973},{"id":"https://openalex.org/keywords/prompt-engineering","display_name":"prompt engineering","score":0.4970000088214874},{"id":"https://openalex.org/keywords/llama-3","display_name":"Llama 3","score":0.6729999780654907},{"id":"https://openalex.org/keywords/gemini-1-5-flash","display_name":"Gemini 1.5 Flash","score":0.9480000138282776}],"concepts":[{"id":"https://openalex.org/C95713431","wikidata":"https://www.wikidata.org/wiki/Q631425","display_name":"Vulnerability (computing)","level":2,"score":0.7684675455093384},{"id":"https://openalex.org/C85641259","wikidata":"https://www.wikidata.org/wiki/Q290042","display_name":"Turn (biochemistry)","level":2,"score":0.5264718532562256},{"id":"https://openalex.org/C143724316","wikidata":"https://www.wikidata.org/wiki/Q312468","display_name":"Series (stratigraphy)","level":2,"score":0.48999515175819397},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.4001748859882355},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.30107176303863525},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.28023630380630493},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.14124727249145508},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.1349879503250122},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.0},{"id":"https://openalex.org/C46141821","wikidata":"https://www.wikidata.org/wiki/Q209402","display_name":"Nuclear magnetic resonance","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.36227/techrxiv.172788963.37426055/v1","doi":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","is_oa":true,"landing_page_url":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","pdf_url":"https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.172788963.37426055","source":null,"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"posted-content"}],"best_oa_location":{"id":"doi:10.36227/techrxiv.172788963.37426055/v1","doi":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","is_oa":true,"landing_page_url":"https://doi.org/10.36227/techrxiv.172788963.37426055/v1","pdf_url":"https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.172788963.37426055","source":null,"license":"cc-by-nc-sa","license_id":"https://openalex.org/licenses/cc-by-nc-sa","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":null,"raw_type":"posted-content"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4403051664.pdf","grobid_xml":"https://content.openalex.org/works/W4403051664.grobid-xml"},"referenced_works_count":4,"referenced_works":["https://openalex.org/W4389518968","https://openalex.org/W4393932428","https://openalex.org/W4402386010","https://openalex.org/W4403620136"],"related_works":["https://openalex.org/W4403620136","https://openalex.org/W4416045849","https://openalex.org/W4416242612","https://openalex.org/W4415971916","https://openalex.org/W4402706410","https://openalex.org/W4416019575","https://openalex.org/W4415337353","https://openalex.org/W7155505014","https://openalex.org/W7110914147","https://openalex.org/W7125766980"],"abstract_inverted_index":{"This":[0],"paper":[1],"builds":[2],"on":[3,6,16,93],"previous":[4],"research":[5],"the":[7,17,38,82,88,91,100,104,110,116,120,127],"Single-Turn":[8],"Crescendo":[9],"Attack":[10],"(STCA)":[11],"by":[12,109],"introducing":[13],"new":[14],"findings":[15],"effectiveness":[18],"of":[19,90,119],"a":[20,145],"four-turn":[21],"STCA":[22],"(STCA-4)":[23],"against":[24],"Open":[25],"AI":[26],"GPT":[27,43,47,94,111],"o1":[28,44,48,112,121],"series":[29,122],"(Strawberry)":[30],"models.":[31,123],"Building":[32],"upon":[33],"our":[34],"original":[35],"work":[36],"with":[37],"STCA-3,":[39],"we":[40],"demonstrate":[41],"that":[42,87,140],"Mini":[45],"and":[46,65,74],"Preview":[49],"models":[50,69,95,113,135],"are":[51],"vulnerable":[52],"to":[53,81,99,136],"an":[54],"STCA-4":[55,92],"attack,":[56],"often":[57],"generating":[58],"disallowed":[59],"content":[60],"such":[61],"as":[62,103],"hate":[63],"speech":[64],"slurs.":[66],"In":[67],"contrast,":[68],"like":[70],"Gemini":[71],"1.5":[72],"Flash":[73],"LLaMA":[75],"3":[76],"8B":[77],"showed":[78],"robust":[79],"resistance":[80],"same":[83],"attack.":[84],"We":[85],"hypothesize":[86],"success":[89],"may":[96],"be":[97],"due":[98],"familiarity":[101],"effect,":[102],"attack":[105],"leverages":[106],"language":[107,134],"generated":[108],"themselves,":[114],"or":[115],"slow-thinking":[117],"ability":[118],"These":[124],"results":[125],"highlight":[126],"need":[128],"for":[129],"improved":[130],"safeguards":[131],"in":[132],"large":[133],"prevent":[137],"adversarial":[138],"exploits":[139],"condense":[141],"multi-turn":[142],"escalations":[143],"into":[144],"single":[146],"prompt.":[147]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2024-10-02T00:00:00"}