Hacker News new | ask | show | jobs
Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents (arxiv.org)
3 points by zhinit 17 days ago