Revealing a Hidden Class of Task-in-Prompt Adversarial Attacks on LLMs

1 points | by aghuang an hour ago

No comments yet