1 paper · 1 filter
Hanqi Li, Jing Peng, Zijian Wang +2
LLM-based multi-agent systems increasingly coordinate planning, reasoning, tool use, and human interaction, yet their reliability remains limited. A central source of this limitati…