https://arxiv.org/abs/2210.03629 ReAct: Synergizing Reasoning and Acting in Language ModelsWhile large language models (LLMs) have demonstrated impressive capabilities across tasks in language understanding and interactive decision making, their abilities for reasoning (e.g. chain-of-thought prompting) and acting (e.g. action plan generation) haarxiv.org Agent 관련 논문에서 빠질 수 없는 것이 ReAct기에...한 번 읽어..