MMLU-Pro Benchmark

What is MMLU-Pro?

MMLU-Pro is the extension of MMLU, integrating more challenging, reasoning-focused questions and expanding the choice set from four to ten options.

Released

2024-06-03

Evaluates

学科, Examination, NeurIPS 2024, 大语言模型, LLM, 知识储备, Knowledge, 不支持, Unsupported

Openness

open

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/MMLU-Pro

MMLU-Pro paper, code and dataset

• Paper https://arxiv.org/abs/2406.01574
• Code https://github.com/TIGER-AI-Lab/MMLU-Pro
• Dataset https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro

Which sources cite MMLU-Pro?

• MMLU-Pro

Related 学科, Examination, NeurIPS 2024, 大语言模型, LLM, 知识储备, Knowledge, 不支持, Unsupported benchmarks

• A-Bench
• A-OKVQA
• AbsPyramid
• ActionAtlas
• Ada-LEval
• AECBench
• AFQMC
• AgentBoard