{"id":"https://openalex.org/W4402671794","doi":"https://doi.org/10.18653/v1/2024.acl-long.422","title":"Text Embedding Inversion Security for Multilingual Language Models","display_name":"Text Embedding Inversion Security for Multilingual Language Models","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W4402671794","doi":"https://doi.org/10.18653/v1/2024.acl-long.422"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2024.acl-long.422","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2024.acl-long.422","pdf_url":"https://aclanthology.org/2024.acl-long.422.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 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2024.acl-long.422.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5045638055","display_name":"Yiyi Chen","orcid":"https://orcid.org/0000-0001-9977-8960"},"institutions":[{"id":"https://openalex.org/I891191580","display_name":"Aalborg University","ror":"https://ror.org/04m5j1k67","country_code":"DK","type":"education","lineage":["https://openalex.org/I891191580"]}],"countries":["DK"],"is_corresponding":false,"raw_author_name":"Yiyi Chen","raw_affiliation_strings":["Department of Computer Science , Aalborg University , Denmark"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Aalborg University , Denmark","institution_ids":["https://openalex.org/I891191580"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069632640","display_name":"Heather Lent","orcid":"https://orcid.org/0000-0003-3463-0767"},"institutions":[{"id":"https://openalex.org/I891191580","display_name":"Aalborg University","ror":"https://ror.org/04m5j1k67","country_code":"DK","type":"education","lineage":["https://openalex.org/I891191580"]}],"countries":["DK"],"is_corresponding":false,"raw_author_name":"Heather Lent","raw_affiliation_strings":["Department of Computer Science , Aalborg University , Denmark"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Aalborg University , Denmark","institution_ids":["https://openalex.org/I891191580"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5013472329","display_name":"Johannes Bjerva","orcid":"https://orcid.org/0000-0002-9512-0739"},"institutions":[{"id":"https://openalex.org/I891191580","display_name":"Aalborg University","ror":"https://ror.org/04m5j1k67","country_code":"DK","type":"education","lineage":["https://openalex.org/I891191580"]}],"countries":["DK"],"is_corresponding":false,"raw_author_name":"Johannes Bjerva","raw_affiliation_strings":["Department of Computer Science , Aalborg University , Denmark"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Aalborg University , Denmark","institution_ids":["https://openalex.org/I891191580"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I891191580"],"apc_list":null,"apc_paid":null,"fwci":2.9858,"has_fulltext":true,"cited_by_count":10,"citation_normalized_percentile":{"value":0.92326063,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"7808","last_page":"7827"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9740999937057495,"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/T10028","display_name":"Topic Modeling","score":0.9740999937057495,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9624000191688538,"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/computer-science","display_name":"Computer science","score":0.7441664338111877},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.661213755607605},{"id":"https://openalex.org/keywords/inversion","display_name":"Inversion (geology)","score":0.5929688215255737},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4860861599445343},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4142014980316162},{"id":"https://openalex.org/keywords/geology","display_name":"Geology","score":0.11572656035423279}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7441664338111877},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.661213755607605},{"id":"https://openalex.org/C1893757","wikidata":"https://www.wikidata.org/wiki/Q3653001","display_name":"Inversion (geology)","level":3,"score":0.5929688215255737},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4860861599445343},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4142014980316162},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.11572656035423279},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.0},{"id":"https://openalex.org/C109007969","wikidata":"https://www.wikidata.org/wiki/Q749565","display_name":"Structural basin","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2024.acl-long.422","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2024.acl-long.422","pdf_url":"https://aclanthology.org/2024.acl-long.422.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 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2024.acl-long.422","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2024.acl-long.422","pdf_url":"https://aclanthology.org/2024.acl-long.422.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 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.5,"display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4402671794.pdf","grobid_xml":"https://content.openalex.org/works/W4402671794.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2081900870","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Textual":[0],"data":[1],"is":[2,130],"often":[3],"represented":[4],"as":[5,21,28,36],"realnumbered":[6],"embeddings":[7,29],"in":[8,106,143],"NLP,":[9],"particularly":[10],"with":[11],"the":[12,50,80,131,141],"popularity":[13],"of":[14,49,82],"large":[15],"language":[16],"models":[17],"(LLMs)":[18],"and":[19,85,89,126,145,149],"Embeddings":[20],"a":[22,118],"Service":[23],"(EaaS).However,":[24],"storing":[25],"sensitive":[26],"information":[27],"can":[30,41],"be":[31,42,100,112],"susceptible":[32],"to":[33,69,103,133],"security":[34,74],"breaches,":[35],"research":[37],"shows":[38],"that":[39,96],"text":[40],"reconstructed":[43],"from":[44],"embeddings,":[45],"even":[46],"without":[47],"knowledge":[48],"underlying":[51],"model.While":[52],"defence":[53],"mechanisms":[54],"have":[55],"been":[56],"explored,":[57],"these":[58],"are":[59],"exclusively":[60],"focused":[61],"on":[62,140],"English,":[63],"leaving":[64],"other":[65],"languages":[66],"potentially":[67],"exposed":[68],"attacks.This":[70],"work":[71],"explores":[72],"LLM":[73],"through":[75],"multilingual":[76,84,97,127,135,150],"embedding":[77],"inversion.We":[78],"define":[79],"problem":[81],"black-box":[83],"crosslingual":[86],"inversion":[87,104,136],"attacks,":[88,105,137],"explore":[90],"their":[91],"potential":[92],"implications.Our":[93],"findings":[94],"suggest":[95],"LLMs":[98],"may":[99,111],"more":[101],"vulnerable":[102],"part":[107],"because":[108],"English-based":[109],"defences":[110],"ineffective.To":[113],"alleviate":[114],"this,":[115],"we":[116],"propose":[117],"simple":[119],"masking":[120],"defense":[121],"effective":[122],"for":[123],"both":[124],"monolingual":[125,148],"models.This":[128],"study":[129],"first":[132],"investigate":[134],"shedding":[138],"light":[139],"differences":[142],"attacks":[144],"defenses":[146],"across":[147],"settings.":[151]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":8},{"year":2024,"cited_by_count":1}],"updated_date":"2026-06-26T08:34:08.712188","created_date":"2025-10-10T00:00:00"}