PhySense: Defending Physically Realizable Attacks for Autonomous Systems via Consistency Reasoning
Published in ACM Conference on Computer and Communications Security (CCS), 2024
This paper is about investigating LLM jailbreak threats, from the perspective of empirical evaluation and automatic generation.
Recommended citation: Z. Yu, A. Li, R. Wen, Y. Chen, N. Zhang. PhySense: Defending Physically Realizable Attacks for Autonomous Systems via Consistency Reasoning. In Proceedings of the 2024 ACM SIGSAC Conference on Computer and Communications Security (CCS) 2024 Oct 14.