1 paper · 1 filter
Zeyang Sha, Hanling Tian, Zhuoer Xu +3
The emergence of autonomous Large Language Model (LLM) agents capable of tool usage has introduced new safety risks that go beyond traditional conversational misuse. These agents,…