1
LACUNA: Safe Agents as Recursive Program Holes
将安全智能体建模为递归程序孔洞，提供一种形式化保障AI行为安全的新思路。
arXiv:2605.28617v1 Announce Type: new Abstract: LLM agents increasingly act by writing code, yet a split persists between the runtime that drives the …
将安全智能体建模为递归程序孔洞，提供一种形式化保障AI行为安全的新思路。
arXiv:2605.28617v1 Announce Type: new Abstract: LLM agents increasingly act by writing code, yet a split persists between the runtime that drives the …