InfoResearchPeer-reviewedLLM-specific
Enhancing Targeted Adversarial Attacks on Large Vision-Language Models via Intermediate Projector
- Published
- Record updated
Summary
Researchers show that black-box targeted attacks on Large Vision-Language Models can be made more precise by working through the projector, a semantic bridge between vision and language. They propose Intermediate Projector Guided Attack (IPGA), which aligns Q-Former query outputs with a target and transfers across models, and IPGA with Residual Query Alignment (IPGA-R), which also preserves non-target content for fine-grained edits. The authors report that IPGA beats baselines on global attacks, IPGA-R wins on fine-grained attacks, and the method transfers to Google Gemini and OpenAI GPT.
Topics
Related items
- InfoSoftBank plunges over 16% as AI sell-off hits Asian tech stocks following Anthropic debutSame vendor · CNBC Technology
- InfoGPT-6.2 Astra vs. Sol: Features, Pricing & Security | StraikerSame vendor · Straiker Blog
- InfoGoogle is putting AI chats right on the homepageSame vendor · The Verge (AI)
- InfoThreatsDay: GPT-6.2 Jailbreak Flaws, BubbleLoader PNG Malware, $500M Ransom Demands + 17 More StoriesSame vendor · The Hacker News
- InfoGemini on Mac can now use your computerSame vendor · The Verge (AI)