{"id":"https://openalex.org/W7163528551","doi":"https://doi.org/10.48550/arxiv.2606.04071","title":"Covert Influence Between Language Models","display_name":"Covert Influence Between Language Models","publication_year":2026,"publication_date":"2026-06-02","ids":{"openalex":"https://openalex.org/W7163528551","doi":"https://doi.org/10.48550/arxiv.2606.04071"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.04071","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.04071","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":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.04071","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5104315293","display_name":"Avidan Shah","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shah, Avidan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137900331","display_name":"Jay Chooi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chooi, Jay","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137909074","display_name":"Jinghua Ou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ou, Jinghua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137911489","display_name":"Shi Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feng, Shi","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":false,"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/T10883","display_name":"Ethics and Social Impacts of AI","score":0.16539999842643738,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.16539999842643738,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.08290000259876251,"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/T11448","display_name":"Face recognition and analysis","score":0.04360000044107437,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/covert","display_name":"Covert","score":0.8985999822616577},{"id":"https://openalex.org/keywords/attribution","display_name":"Attribution","score":0.5573999881744385},{"id":"https://openalex.org/keywords/phenomenon","display_name":"Phenomenon","score":0.5558000206947327},{"id":"https://openalex.org/keywords/payload","display_name":"Payload (computing)","score":0.5436000227928162},{"id":"https://openalex.org/keywords/pointwise","display_name":"Pointwise","score":0.5008999705314636},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.37389999628067017},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.3653999865055084}],"concepts":[{"id":"https://openalex.org/C2779338814","wikidata":"https://www.wikidata.org/wiki/Q5179285","display_name":"Covert","level":2,"score":0.8985999822616577},{"id":"https://openalex.org/C143299363","wikidata":"https://www.wikidata.org/wiki/Q900584","display_name":"Attribution","level":2,"score":0.5573999881744385},{"id":"https://openalex.org/C50335755","wikidata":"https://www.wikidata.org/wiki/Q483247","display_name":"Phenomenon","level":2,"score":0.5558000206947327},{"id":"https://openalex.org/C134066672","wikidata":"https://www.wikidata.org/wiki/Q1424639","display_name":"Payload (computing)","level":3,"score":0.5436000227928162},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5236999988555908},{"id":"https://openalex.org/C2777984123","wikidata":"https://www.wikidata.org/wiki/Q9248237","display_name":"Pointwise","level":2,"score":0.5008999705314636},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.44929999113082886},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.39579999446868896},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.37389999628067017},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.3653999865055084},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.3321000039577484},{"id":"https://openalex.org/C29024540","wikidata":"https://www.wikidata.org/wiki/Q1476964","display_name":"Covert channel","level":5,"score":0.32089999318122864},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.29260000586509705},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.288100004196167},{"id":"https://openalex.org/C64357122","wikidata":"https://www.wikidata.org/wiki/Q1149766","display_name":"Causality (physics)","level":2,"score":0.2867000102996826},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.2847999930381775},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.2793000042438507},{"id":"https://openalex.org/C3017944768","wikidata":"https://www.wikidata.org/wiki/Q1450463","display_name":"Poison control","level":2,"score":0.26739999651908875},{"id":"https://openalex.org/C46312422","wikidata":"https://www.wikidata.org/wiki/Q11024","display_name":"Communication","level":1,"score":0.25839999318122864},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.2574999928474426},{"id":"https://openalex.org/C194995250","wikidata":"https://www.wikidata.org/wiki/Q531136","display_name":"Affordance","level":2,"score":0.25189998745918274}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.04071","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.04071","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":"doi:10.48550/arxiv.2606.04071","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.04071","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":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"As":[0],"language":[1],"models":[2],"increasingly":[3],"consume":[4],"one":[5],"another's":[6],"outputs,":[7],"covert":[8,76,106,145],"influence":[9,62,77,107,146],"--":[10,34],"a":[11,14,27,36,112,160],"phenomenon":[12,114],"where":[13],"sender's":[15],"payload":[16,93],"(the":[17],"behavioral":[18],"disposition":[19],"it":[20],"is":[21,111,124,147],"conditioned":[22],"to":[23,26,85,127,162],"propagate)":[24],"transfers":[25,94],"receiver":[28],"through":[29],"carriers":[30,87,110],"undetectable":[31],"by":[32],"humans":[33],"becomes":[35],"growing":[37],"risk.":[38],"We":[39,101],"characterize":[40],"this":[41],"risk":[42,142],"across":[43,78,133],"three":[44,80],"interfaces:":[45],"supervised":[46],"fine-tuning,":[47],"on-policy":[48],"distillation,":[49],"and":[50,53,130,152,164],"in-context":[51],"learning,":[52],"find":[54],"that":[55,88,95,105,140],"they":[56],"vary":[57],"in":[58],"the":[59,83,122,141],"scale":[60],"of":[61],"achievable":[63],"without":[64],"leaving":[65],"behind":[66],"human-visible":[67],"traces.":[68],"Using":[69],"inference-time":[70],"per-sample":[71],"attribution":[72,156],"scores,":[73],"we":[74,153],"study":[75,154],"all":[79],"interfaces":[81],"with":[82,108],"ability":[84],"select":[86],"amplify":[89],"training-time":[90],"influence,":[91],"unlocking":[92],"prior":[96,116],"work":[97],"could":[98],"not":[99],"achieve.":[100],"further":[102],"provide":[103],"evidence":[104],"natural-language":[109],"distinct":[113],"from":[115],"studies":[117],"using":[118],"number":[119],"carriers,":[120],"as":[121,159],"latter":[123],"more":[125],"resistant":[126],"human":[128],"detection":[129],"less":[131],"portable":[132],"model":[134],"families.":[135],"Together,":[136],"these":[137],"results":[138],"suggest":[139],"surface":[143],"for":[144],"broader":[148],"than":[149],"previously":[150],"recognized,":[151],"pointwise":[155],"scoring":[157],"methods":[158],"tool":[161],"investigate":[163],"mitigate":[165],"it.":[166]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-05T00:00:00"}