Large language model (LLM) agents increasingly operate in open-world environments where safety risks cannot be fully predefined. However, most existing guard…
机构:北航
来源:arXiv 2611.02516 | AI4Papers 论文推荐平台