Fetching the paper…

PLeak: Prompt Leaking Attacks against Large Language Model Applications · Around