{"id":"https://openalex.org/W7161829171","doi":"https://doi.org/10.48550/arxiv.2605.18988","title":"Surviving the Unseen: Predictive Defense for Novel Multi-Turn Multimodal Attacks","display_name":"Surviving the Unseen: Predictive Defense for Novel Multi-Turn Multimodal Attacks","publication_year":2026,"publication_date":"2026-05-18","ids":{"openalex":"https://openalex.org/W7161829171","doi":"https://doi.org/10.48550/arxiv.2605.18988"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.18988","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18988","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"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.2605.18988","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120844795","display_name":"Doohee You","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"You, Doohee","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5120844795"],"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.4945000112056732,"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":0.4945000112056732,"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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.10719999670982361,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.050599999725818634,"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/trajectory","display_name":"Trajectory","score":0.5870000123977661},{"id":"https://openalex.org/keywords/anomaly-detection","display_name":"Anomaly detection","score":0.5805000066757202},{"id":"https://openalex.org/keywords/covariance","display_name":"Covariance","score":0.4765999913215637},{"id":"https://openalex.org/keywords/mahalanobis-distance","display_name":"Mahalanobis distance","score":0.46630001068115234},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.42340001463890076},{"id":"https://openalex.org/keywords/markov-chain","display_name":"Markov chain","score":0.41999998688697815},{"id":"https://openalex.org/keywords/acceleration","display_name":"Acceleration","score":0.40549999475479126},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.351500004529953}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6093999743461609},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.5870000123977661},{"id":"https://openalex.org/C739882","wikidata":"https://www.wikidata.org/wiki/Q3560506","display_name":"Anomaly detection","level":2,"score":0.5805000066757202},{"id":"https://openalex.org/C178650346","wikidata":"https://www.wikidata.org/wiki/Q201984","display_name":"Covariance","level":2,"score":0.4765999913215637},{"id":"https://openalex.org/C1921717","wikidata":"https://www.wikidata.org/wiki/Q1334846","display_name":"Mahalanobis distance","level":2,"score":0.46630001068115234},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.42340001463890076},{"id":"https://openalex.org/C98763669","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov chain","level":2,"score":0.41999998688697815},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.40549999475479126},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.351500004529953},{"id":"https://openalex.org/C39920418","wikidata":"https://www.wikidata.org/wiki/Q11476","display_name":"Kinematics","level":2,"score":0.3479999899864197},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3472999930381775},{"id":"https://openalex.org/C35525427","wikidata":"https://www.wikidata.org/wiki/Q745881","display_name":"Intrusion detection system","level":2,"score":0.3303999900817871},{"id":"https://openalex.org/C79337645","wikidata":"https://www.wikidata.org/wiki/Q779824","display_name":"Outlier","level":2,"score":0.32829999923706055},{"id":"https://openalex.org/C159886148","wikidata":"https://www.wikidata.org/wiki/Q176645","display_name":"Markov process","level":2,"score":0.323199987411499},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3230000138282776},{"id":"https://openalex.org/C34388435","wikidata":"https://www.wikidata.org/wiki/Q2267362","display_name":"Bounded function","level":2,"score":0.3140999972820282},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.311599999666214},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.296099990606308},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2818000018596649},{"id":"https://openalex.org/C163836022","wikidata":"https://www.wikidata.org/wiki/Q6771326","display_name":"Markov model","level":3,"score":0.27799999713897705},{"id":"https://openalex.org/C2775941552","wikidata":"https://www.wikidata.org/wiki/Q25212305","display_name":"Isolation (microbiology)","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C49261128","wikidata":"https://www.wikidata.org/wiki/Q1132455","display_name":"Hazard","level":2,"score":0.25859999656677246},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.2563999891281128},{"id":"https://openalex.org/C111065885","wikidata":"https://www.wikidata.org/wiki/Q1189053","display_name":"Fuzz testing","level":3,"score":0.2524000108242035}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.18988","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18988","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.18988","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.18988","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","score":0.5452724695205688,"display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"The":[0,79,102],"expansion":[1],"of":[2],"Multimodal":[3],"Large":[4],"Language":[5],"Models":[6],"(MLLMs)":[7],"and":[8,54,75,94,124,139,189],"their":[9],"integration":[10],"into":[11,144],"autonomous":[12],"agentic":[13,194],"workflows":[14],"has":[15],"introduced":[16],"a":[17,71,88,99,112,145,152,168,185,198],"non-stationary":[18],"attack":[19],"surface.":[20],"Empirical":[21],"observations":[22],"indicate":[23],"that":[24,30,91,163,177],"adversaries":[25],"employ":[26],"progressive,":[27],"cross-modal":[28],"perturbations":[29],"evade":[31],"turn-specific":[32],"guardrails":[33],"by":[34,46],"distributing":[35],"malicious":[36,135,178],"intent":[37],"across":[38],"longitudinal":[39],"conversational":[40,96],"trajectories.":[41],"Static":[42],"defense":[43],"mechanisms,":[44],"constrained":[45],"the":[47,164],"Markov":[48,155],"property,":[49],"evaluate":[50],"inputs":[51],"in":[52,121],"isolation":[53],"fail":[55],"to":[56,108,117,128],"detect":[57],"cumulative":[58],"structural":[59,105],"poisoning.":[60],"To":[61],"handle":[62],"this":[63,65],"limitation,":[64],"paper":[66],"formulates":[67],"safety":[68,203],"verification":[69],"as":[70,87,98],"dynamic":[72],"survival":[73],"prediction":[74],"trajectory":[76,126],"dynamics":[77],"problem.":[78],"Triple-tier":[80],"Anomaly":[81],"Defense":[82],"(TRIAD)":[83],"framework":[84,103,166,183],"is":[85],"proposed":[86],"predictive":[89,190],"model":[90,150],"maps":[92],"multimodal":[93],"multi-turn":[95],"flow":[97],"continuous":[100,134,202],"trajectory.":[101],"integrates":[104],"anomaly":[106],"detection":[107],"monitor":[109,118],"covariance":[110,119],"shifts,":[111],"Ledoit-Wolf":[113],"regularized":[114],"Mahalanobis":[115],"distance":[116],"shifts":[120],"high-dimensional":[122],"spaces,":[123],"topological":[125],"acceleration":[127,179],"differentiate":[129],"benign":[130],"creative":[131],"exploration":[132],"from":[133],"drift.":[136],"These":[137],"kinematic":[138],"geometric":[140],"features":[141],"are":[142],"integrated":[143],"time-varying":[146],"Cox":[147],"Proportional":[148],"Hazards":[149],"via":[151],"Bayesian":[153],"Hidden":[154],"Model":[156],"(HMM)":[157],"feedback":[158],"loop.":[159],"Theoretical":[160],"analysis":[161],"demonstrates":[162],"TRIAD":[165],"provides":[167,184],"mathematically":[169],"bounded":[170],"expected":[171],"time-to-failure":[172],"under":[173],"adversarial":[174],"perturbations,":[175],"ensuring":[176],"diverges":[180],"positively.":[181],"This":[182],"computationally":[186],"efficient,":[187],"interpretable,":[188],"safeguard":[191],"for":[192,201],"real-time":[193],"AI":[195],"systems,":[196],"establishing":[197],"rigorous":[199],"foundation":[200],"alignment":[204],"without":[205],"relying":[206],"on":[207],"empirical":[208],"retraining.":[209]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-21T00:00:00"}