The rapid proliferation of benchmarks for evaluating large language models (LLMs) has created an urgent need for systematic methods to assess benchmark quality itself. We propose Benchmark^2, a comprehensive framework comprising three complementary metrics: (1) Cross-Benchmark Ranking Consistency, measuring whether a benchmark produces model rankings aligned with peer benchmarks; (2) Discriminability Score, quantifying a benchmark's ability to differentiate between models; and (3) Capability Alignment Deviation, identifying problematic instances where stronger models fail but weaker models succeed within the same model family. We conduct extensive experiments across 15 benchmarks spanning mathematics, reasoning, and knowledge domains, evaluating 11 LLMs across four model families. Our analysis reveals significant quality variations among existing benchmarks and demonstrates that selective benchmark construction based on our metrics can achieve comparable evaluation performance with substantially reduced test sets.
@article{arxiv.2601.03986,
title = {Benchmark^2: Systematic Evaluation of LLM Benchmarks},
author = {Qi Qian and Chengsong Huang and Jingwen Xu and Changze Lv and Muling Wu and Wenhao Liu and Xiaohua Wang and Zhenghua Wang and Zisu Huang and Muzhao Tian and Jianhan Xu and Kun Hu and He-Da Wang and Yao Hu and Xuanjing Huang and Xiaoqing Zheng},
journal= {arXiv preprint arXiv:2601.03986},
year = {2026}
}