{"data":{"id":"e9e99741-2dc2-4d35-b4d7-4ea9e98f3bec","title":"Where Is Rank-1 DiM Refusal Intervenable in Multimodal LM Derivatives?","summary":"This research asks whether causal chat attack lift under a shared mean / rank-1 difference-in-means (DiM) protocol concentrates at the text decoder T or also appears at perceptual encoders, projectors and fusion paths (E/P/F). On HADES adversarial images paired with harmful text and spoken AdvBench, text DiM at T raises attack success rate (ASR) while the same DiM at E/P/F does not. On VLGuard Unsafe prompts (n=200, Llama-Guard) the same T vs. P/E directionality holds, with ASR rising from 0.030 to 0.150 at T.","solution":"N/A -- no mitigation discussed in source.","labels":["research","safety"],"sourceUrl":"https://arxiv.org/abs/2611.19826v1","publishedAt":"2026-11-15T06:15:45.000Z","cveId":null,"cweIds":null,"cvssScore":null,"cvssSeverity":null,"severity":"info","attackType":["jailbreak"],"issueType":"research","affectedPackages":null,"affectedPackageNames":null,"affectedPackageRefs":null,"affectedVendors":[],"affectedVendorsRaw":["Llama-Guard","VLGuard","HADES","AdvBench"],"classifierModel":"claude-haiku-5-5","classifierPromptVersion":"v4","summaryPromptVersion":"v2","headline":null,"headlinePromptVersion":null,"cvssVector":null,"attackVector":null,"attackComplexity":null,"privilegesRequired":null,"userInteraction":null,"exploitMaturity":null,"epssScore":null,"epssCheckedAt":null,"kevDateAdded":null,"advisoryAliases":null,"affectedPackagesSource":null,"affectedPackagesCheckedAt":null,"patchAvailable":null,"disclosureDate":"2026-11-15T06:15:45.000Z","fixReleasedAt":null,"daysToFix":null,"fixCheckedAt":null,"capecIds":null,"crossRefCount":0,"attackSophistication":"advanced","impactType":["safety"],"aiComponentTargeted":"model","llmSpecific":false,"classifierConfidence":0.9,"researchCategory":"preprint","atlasIds":["AML.T0054"]}}