论文地址:ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate | OpenReviewText evaluation has historically posed significant challenges, often demanding substantial labor and time cost. With the emergence of large language models (LLMs…