Zhikun Zhang

Results show that apparent attack winners depend on text-quality constraints, attack generality, and capability assumptions, and that apparent attack winners depend on text-quality constraints, attack generality, and capability assumptions.

This paper introduces InceptionRAG, a stealthy attack mechanism that subverts the standard attack paradigm of RAG, and shows superior evasion capabilities, effectively bypassing established defenses that mitigate traditional single-document injections.

Multi-item users can face significantly more powerful poisoning attacks, but a new method achieves robust estimation without sacrificing performance.