MMLU Chat

Chat-format variant of the Massive Multitask Language Understanding benchmark, evaluating language models across 57 tasks including elementary mathematics, US history, computer science, law, and other professional and academic subjects. This version uses conversational prompting format for model evaluation.

Released

Unknown

Readiness

Paper only

Primary field

Finance & Economics

Benchmark Radar records only publicly supported details and links back to primary sources for verification.