Fetching the paper…

SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks · Around