MMLU Reasoning Benchmark

What is MMLU Reasoning Benchmark?

The MMLU Reasoning Benchmark comprises multiple-choice questions spanning subjects such as mathematics, history, computer science, and law. This benchmark evaluates a Large Language Model’s (LLM) ability to showcase knowledge and comprehension across diverse academic disciplines.

Resources:

• MMLU dataset: https://github.com/hendrycks/test
• MMLU Paper: https://arxiv.org/abs/2009.03300