MiniLongBench

What is MiniLongBench?

MiniLongBench is a low-cost benchmark for evaluating the Long Context Understanding (LCU) capabilities of LLMs, featuring a compact yet diverse test set of…

Released

2025-05-15

Evaluates

理解, Understanding, 语言, Language, 长文本, Long-Context, 大语言模型, LLM, 语言理解, Comprehension, 长上下文, Long Context, 不支持, Unsupported

Openness

open

Importer

OpenCompass Hub

Reported scores

0

Source provenance

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

MiniLongBench paper, code and dataset

• Paper https://arxiv.org/abs/2505.19959
• Code https://github.com/MilkThink-Lab/MiniLongBench
• Dataset https://huggingface.co/datasets/linggm/MiniLongBench

Which sources cite MiniLongBench?

• MiniLongBench

Related 理解, Understanding, 语言, Language, 长文本, Long-Context, 大语言模型, LLM, 语言理解, Comprehension, 长上下文, Long Context, 不支持, Unsupported benchmarks

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