MIB Benchmark

What is MIB?

MIB is a benchmark designed to evaluate mechanistic interpretability methods for neural language models. Benchmark Radar tracks MIB, cited by 1 source…

Released

2025-04-17

Evaluates

理解, Understanding, 大语言模型, LLM, 语言理解, Comprehension, 不支持, Unsupported

Openness

unknown

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

MIB paper and code

• Paper https://arxiv.org/abs/2407.14494
• Paper https://arxiv.org/abs/2504.13151v1
• Code https://github.com/aaronmueller/mib

Which sources cite MIB?

• MIB

Related 理解, Understanding, 大语言模型, LLM, 语言理解, Comprehension, 不支持, Unsupported benchmarks

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