1 paper
Evan Chen, Shiqiang Wang, Christopher G. Brinton
Distributed LLM-agent teams can read the latest shared facts and still act on an obsolete plan. A planner may derive an action from requirement r3, another agent may commit $r_4…