0%

Agent = Large Language Model (LLM) + Observation + Thought + Action + Memory

MultiAgent = Agents + Environment + Standard Operating Procedure (SOP) + Communication + Economy

ReAct1(reason, act) 是一种使用自然语言推理解决复杂任务的语言模型范例,ReAct旨在用于允许LLM执行某些操作的任务


reference