llm-dra.github.io

Demo site of the paper Making Them Ask and Answer: Jailbreaking Large Language Models in Few Queries via Disguise and Reconstruction