Reflections and Fragments: Securing LLMs Against Sequential Mosaic Attacks
Self-play red-teaming improves language-model safety by pitting attacker and defender roles against each other in a zero-sum game. However, real adversaries increasingly use mosaic attacks: multi-turn sequences whose individual fragments are innocuous in isolation yet assemble into a harmful payload. We develop a theor...
Emanuele La Malfa, Saar Cohen, G. La Malfa et al.
· 0 citations