这是复旦大学自然语言处理实验室设计的专业评估基准，旨在衡量大型语言模型的性能和可靠性。

MOSS is a sophisticated conversational large language model developed by a research team at 复旦大学, designed for high-quality human-AI interaction.