GPQA Benchmark

What is GPQA?

GPQA, a challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry.

Released

2023-11-20

Evaluates

知识, Knowledge, 大语言模型, LLM, 知识储备, 开源收录, Open-Source, 不支持, Unsupported

Openness

restricted

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

GPQA paper, code and dataset

• Paper https://arxiv.org/pdf/2311.12022
• Code https://github.com/idavidrein/gpqa
• Dataset https://huggingface.co/datasets/Idavidrein/gpqa

Which sources cite GPQA?

• GPQA

Related 知识, Knowledge, 大语言模型, LLM, 知识储备, 开源收录, Open-Source, 不支持, Unsupported benchmarks

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