LLM agents are increasingly deployed in Multi-Agent Systems (MAS), and they introduce the risk of coordination among agents that might escape human oversight…
机构:LSE
来源:arXiv 2611.02153 | AI4Papers 论文推荐平台