CEB Benchmark

What is CEB?

CEB evaluates LLM bias compositionally, featuring 11k samples characterized across bias types, social groups, and tasks.

Released

2024-07-03

Evaluates

其他, Other, 大语言模型, LLM, 安全对齐, Safety Alignment, 不支持, Unsupported

Openness

restricted

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

CEB paper, code and dataset

• Paper https://arxiv.org/abs/2407.02408
• Code https://github.com/SongW-SW/CEB
• Dataset https://huggingface.co/datasets/Song-SW/CEB

Which sources cite CEB?

• CEB

Related 其他, Other, 大语言模型, LLM, 安全对齐, Safety Alignment, 不支持, Unsupported benchmarks

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