{"id":"https://openalex.org/W4417529639","doi":"https://doi.org/10.48550/arxiv.2512.16059","title":"ContextLeak: Auditing Leakage in Private In-Context Learning Methods","display_name":"ContextLeak: Auditing Leakage in Private In-Context Learning Methods","publication_year":2025,"publication_date":"2025-12-18","ids":{"openalex":"https://openalex.org/W4417529639","doi":"https://doi.org/10.48550/arxiv.2512.16059"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2512.16059","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2512.16059","pdf_url":"https://arxiv.org/pdf/2512.16059","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2512.16059","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5113123616","display_name":"Jacob Choi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Choi, Jacob","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102874627","display_name":"Shougen Cao","orcid":"https://orcid.org/0000-0002-0858-8940"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cao, Shuying","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Dong, Xingjian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dong, Xingjian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Banayeeanzade, Amin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Banayeeanzade, Amin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060852587","display_name":"W. J. Zhu","orcid":"https://orcid.org/0000-0003-2618-0436"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Wang Bill","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041906762","display_name":"Robin Jia","orcid":"https://orcid.org/0009-0002-8123-7132"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jia, Robin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5120869636","display_name":"Sai Praneeth Karimireddy","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Karimireddy, Sai Praneeth","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":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.6442999839782715,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.6442999839782715,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.09989999979734421,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.04560000076889992,"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/audit","display_name":"Audit","score":0.6934999823570251},{"id":"https://openalex.org/keywords/leakage","display_name":"Leakage (economics)","score":0.6410999894142151},{"id":"https://openalex.org/keywords/information-leakage","display_name":"Information leakage","score":0.6072999835014343},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.5306000113487244},{"id":"https://openalex.org/keywords/private-information-retrieval","display_name":"Private information retrieval","score":0.4478999972343445},{"id":"https://openalex.org/keywords/information-sensitivity","display_name":"Information sensitivity","score":0.44519999623298645}],"concepts":[{"id":"https://openalex.org/C199521495","wikidata":"https://www.wikidata.org/wiki/Q181487","display_name":"Audit","level":2,"score":0.6934999823570251},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.66839998960495},{"id":"https://openalex.org/C2777042071","wikidata":"https://www.wikidata.org/wiki/Q6509304","display_name":"Leakage (economics)","level":2,"score":0.6410999894142151},{"id":"https://openalex.org/C2779201187","wikidata":"https://www.wikidata.org/wiki/Q2775060","display_name":"Information leakage","level":2,"score":0.6072999835014343},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.5306000113487244},{"id":"https://openalex.org/C99221444","wikidata":"https://www.wikidata.org/wiki/Q1532069","display_name":"Private information retrieval","level":2,"score":0.4478999972343445},{"id":"https://openalex.org/C137822555","wikidata":"https://www.wikidata.org/wiki/Q2587068","display_name":"Information sensitivity","level":2,"score":0.44519999623298645},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.4327999949455261},{"id":"https://openalex.org/C2776889888","wikidata":"https://www.wikidata.org/wiki/Q1135789","display_name":"Unintended consequences","level":2,"score":0.429500013589859},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.39980000257492065},{"id":"https://openalex.org/C71745522","wikidata":"https://www.wikidata.org/wiki/Q2476929","display_name":"Confidentiality","level":2,"score":0.39320001006126404},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3856000006198883},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36169999837875366},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3587999939918518},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.35269999504089355},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3474000096321106},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.30570000410079956},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.2621999979019165},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.25110000371932983}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2512.16059","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2512.16059","pdf_url":"https://arxiv.org/pdf/2512.16059","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2512.16059","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2512.16059","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2512.16059","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2512.16059","pdf_url":"https://arxiv.org/pdf/2512.16059","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4417529639.pdf","grobid_xml":"https://content.openalex.org/works/W4417529639.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"In-Context":[0],"Learning":[1],"(ICL)":[2],"has":[3],"become":[4],"a":[5,107,145],"standard":[6],"technique":[7],"for":[8],"adapting":[9],"Large":[10],"Language":[11],"Models":[12],"(LLMs)":[13],"to":[14,36,49,64,74,99],"specialized":[15],"tasks":[16],"by":[17],"supplying":[18],"task-specific":[19],"exemplars":[20,27],"within":[21],"the":[22,71,77,92,135,140],"prompt.":[23],"However,":[24],"when":[25],"these":[26,66],"contain":[28],"sensitive":[29,93,166],"information,":[30],"reliable":[31],"privacy-preserving":[32,44],"mechanisms":[33],"are":[34,59],"essential":[35],"prevent":[37],"unintended":[38],"leakage":[39,53,80,131,136],"through":[40],"model":[41],"outputs.":[42],"Many":[43],"methods":[45,121,158],"have":[46],"been":[47],"proposed":[48],"protect":[50],"against":[51],"information":[52,79,167],"in":[54,81,91],"this":[55],"context,":[56],"but":[57],"there":[58],"fewer":[60],"efforts":[61],"on":[62],"how":[63],"audit":[65],"methods.":[67],"We":[68,103,125],"introduce":[69],"ContextLeak,":[70],"first":[72],"framework":[73],"empirically":[75],"measure":[76],"worst-case":[78,149],"ICL.":[82],"ContextLeak":[83,105,128],"uses":[84],"canary":[85],"insertion,":[86],"embedding":[87],"uniquely":[88],"identifiable":[89],"tokens":[90],"dataset":[94],"and":[95,118,134],"crafting":[96],"targeted":[97],"queries":[98],"detect":[100],"their":[101],"presence.":[102],"apply":[104],"across":[106,132],"range":[108],"of":[109,148],"private":[110,120],"ICL":[111],"techniques,":[112],"including":[113],"both":[114],"heuristic":[115],"prompt-based":[116],"defenses":[117],"differentially":[119],"with":[122,139],"formal":[123],"guarantees.":[124],"show":[126],"that":[127,156],"reliably":[129],"detects":[130],"methods,":[133],"increases":[137],"monotonically":[138],"theoretical":[141],"privacy":[142,150],"budget,":[143],"offering":[144],"practical":[146],"signal":[147],"risk.":[151],"Our":[152],"analysis":[153],"further":[154],"reveals":[155],"existing":[157],"strike":[159],"poor":[160],"privacy-utility":[161],"trade-offs,":[162],"either":[163],"completely":[164],"leaking":[165],"or":[168],"severely":[169],"degrading":[170],"performance.":[171]},"counts_by_year":[],"updated_date":"2026-07-18T07:39:51.176621","created_date":"2025-12-21T00:00:00"}