Explore high-quality datasets for your AI and machine learning projects.
CMMLU is a comprehensive Chinese evaluation suite specifically designed to assess large‑scale multi‑task language understanding capabilities in Chinese linguistic and cultural contexts. It covers 67 subjects ranging from basic to advanced professional levels, including STEM fields such as physics and mathematics as well as humanities and social sciences. Many tasks involve nuanced phrasing and cultural specifics that are hard to translate. Answers for many tasks are China‑specific and may not be applicable elsewhere. Each subject provides development and test sets; every question is a four‑option multiple‑choice item with a single correct answer.
