MASK Benchmark

What is MASK?

The MASK evaluation provides a rigorous benchmark for evaluating honesty in large language models by measuring whether models remain truthful when incentivized…

Released

2025-03-05

Evaluates

其他, Other, 大语言模型, LLM, 安全对齐, Safety Alignment, 事实可靠性, Factual Reliability, 不支持, Unsupported

Openness

restricted

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

MASK paper, code and dataset

• Paper https://arxiv.org/abs/2503.03750
• Code https://github.com/centerforaisafety/mask
• Dataset https://huggingface.co/datasets/cais/MASK

Which sources cite MASK?

• MASK

Related 其他, Other, 大语言模型, LLM, 安全对齐, Safety Alignment, 事实可靠性, Factual Reliability, 不支持, Unsupported benchmarks

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