{"id":"https://openalex.org/W4416035174","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.242","title":"Backdoor-Powered Prompt Injection Attacks Nullify Defense Methods","display_name":"Backdoor-Powered Prompt Injection Attacks Nullify Defense Methods","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416035174","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.242"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.242","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.242","pdf_url":"https://aclanthology.org/2025.findings-emnlp.242.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: EMNLP 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.242.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100398890","display_name":"Yulin Chen","orcid":"https://orcid.org/0000-0001-5679-4055"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yulin Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086448068","display_name":"Haoran Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haoran Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5119204421","display_name":"Yuan Sui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuan Sui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020880385","display_name":"Yangqiu Song","orcid":"https://orcid.org/0000-0002-7818-6090"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yangqiu Song","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5065675832","display_name":"Bryan Hooi","orcid":"https://orcid.org/0000-0002-5645-1754"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bryan Hooi","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":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"4508","last_page":"4527"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.3709000051021576,"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":0.3709000051021576,"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":0.25040000677108765,"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/T12479","display_name":"Web Application Security Vulnerabilities","score":0.08340000361204147,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.9754999876022339},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.5001000165939331},{"id":"https://openalex.org/keywords/hierarchy","display_name":"Hierarchy","score":0.4602000117301941},{"id":"https://openalex.org/keywords/virtuous-circle-and-vicious-circle","display_name":"Virtuous circle and vicious circle","score":0.39469999074935913},{"id":"https://openalex.org/keywords/sophistication","display_name":"Sophistication","score":0.2989000082015991}],"concepts":[{"id":"https://openalex.org/C2781045450","wikidata":"https://www.wikidata.org/wiki/Q254569","display_name":"Backdoor","level":2,"score":0.9754999876022339},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.6363000273704529},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5803999900817871},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.5001000165939331},{"id":"https://openalex.org/C31170391","wikidata":"https://www.wikidata.org/wiki/Q188619","display_name":"Hierarchy","level":2,"score":0.4602000117301941},{"id":"https://openalex.org/C23979983","wikidata":"https://www.wikidata.org/wiki/Q1165103","display_name":"Virtuous circle and vicious circle","level":2,"score":0.39469999074935913},{"id":"https://openalex.org/C168725872","wikidata":"https://www.wikidata.org/wiki/Q991663","display_name":"Sophistication","level":2,"score":0.2989000082015991},{"id":"https://openalex.org/C95713431","wikidata":"https://www.wikidata.org/wiki/Q631425","display_name":"Vulnerability (computing)","level":2,"score":0.2635999917984009},{"id":"https://openalex.org/C17623918","wikidata":"https://www.wikidata.org/wiki/Q441828","display_name":"Secondary air injection","level":2,"score":0.259799987077713},{"id":"https://openalex.org/C126965237","wikidata":"https://www.wikidata.org/wiki/Q816335","display_name":"Accidental","level":2,"score":0.2538999915122986}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.242","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.242","pdf_url":"https://aclanthology.org/2025.findings-emnlp.242.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: EMNLP 2025","raw_type":"proceedings-article"},{"id":"pmh:oai:repository.hkust.edu.hk:1783.1-169066","is_oa":false,"landing_page_url":"http://repository.hkust.edu.hk/ir/Record/1783.1-169066","pdf_url":null,"source":{"id":"https://openalex.org/S4306401796","display_name":"Rare & Special e-Zone (The Hong Kong University of Science and Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I200769079","host_organization_name":"Hong Kong University of Science and Technology","host_organization_lineage":["https://openalex.org/I200769079"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Conference paper"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.242","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.242","pdf_url":"https://aclanthology.org/2025.findings-emnlp.242.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: EMNLP 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416035174.pdf","grobid_xml":"https://content.openalex.org/works/W4416035174.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"With":[0],"the":[1,11,20,28,31,41,51,56,62,91,97,105,108,116,119,125,128,130,134,138,143,173],"development":[2],"of":[3,19],"technology,":[4],"large":[5],"language":[6,14],"models":[7],"(LLMs)":[8],"have":[9],"dominated":[10],"downstream":[12],"natural":[13],"processing":[15],"(NLP)":[16],"tasks.However,":[17],"because":[18],"LLMs'":[21],"instruction-following":[22],"abilities":[23],"and":[24,60,123],"inability":[25],"to":[26,45,73],"distinguish":[27],"instructions":[29],"in":[30],"data":[32],"content,":[33],"such":[34],"as":[35],"web":[36],"pages":[37],"from":[38,55],"search":[39],"engines,":[40],"LLMs":[42,52],"are":[43,71,158],"vulnerable":[44],"prompt":[46,77,92,101,112,155,163,168],"injection":[47,78,93,102,113,156,164,169],"attacks.These":[48],"attacks":[49,79,88,157],"trick":[50],"into":[53,127],"deviating":[54],"original":[57],"input":[58],"instruction":[59,67,98,140,174],"executing":[61],"attackers'":[63],"target":[64],"instruction.Recently,":[65],"various":[66],"hierarchy":[68,175],"defense":[69,94,170],"strategies":[70],"proposed":[72],"effectively":[74],"defend":[75],"against":[76],"via":[80],"fine-tuning.In":[81],"this":[82],"paper,":[83],"we":[84],"explore":[85],"more":[86,159],"vicious":[87],"that":[89,153],"nullify":[90],"methods,":[95,171],"even":[96,172],"hierarchy:":[99],"backdoorpowered":[100],"attacks,":[103,165],"where":[104],"attackers":[106,117],"utilize":[107],"backdoor":[109,126],"attack":[110,114],"for":[111,148],"purposes.Specifically,":[115],"poison":[118],"supervised":[120],"fine-tuning":[121],"samples":[122],"insert":[124],"model.Once":[129],"trigger":[131],"is":[132],"activated,":[133],"backdoored":[135],"model":[136],"executes":[137],"injected":[139],"surrounded":[141],"by":[142],"trigger.We":[144],"construct":[145],"a":[146],"benchmark":[147],"comprehensive":[149],"evaluation.Our":[150],"experiments":[151],"demonstrate":[152],"backdoor-powered":[154],"harmful":[160],"than":[161],"previous":[162],"nullifying":[166],"existing":[167],"techniques.":[176],"1":[177]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-11-08T00:00:00"}