Are Benchmarks Reliable? Toward Structural Diagnosis via Sample-Level Capability Boundaries
TL;DR
Experiments show that benchmark reliability is axis-conditioned and heterogeneous: GSM8K and MATH exhibit the most stable measurement structures, MMLU and TriviaQA are relatively stable but heterogeneous, while GPQA and PopQA show stronger axis-conditioned risks.
Abstract
Evaluating large language models (LLMs) relies heavily on benchmark scores, yet aggregate metrics can obscure whether benchmark samples reliably support model comparison. We introduce \textbf{BSDProbe}, a sample-level framework for \emph{benchmark structural diagnosis} that estimates capability boundaries from repeated-response trajectories along ordered model axes. BSDProbe summarizes samples by boundary position, boundary width, boundary-signal validity, and order consistency, then aggregates them into benchmark-level structural profiles. Experiments on six benchmarks show that benchmark reliability is axis-conditioned and heterogeneous: GSM8K and MATH exhibit the most stable measurement structures, MMLU and TriviaQA are relatively stable but heterogeneous, while GPQA and PopQA show stronger axis-conditioned risks. These profiles remain consistent across Qwen3, Qwen2.5, and cross-model axes. BSDProbe further selects compact high-value subsets whose model discriminability reaches up to $8.58\times$ that of the full benchmark. These results suggest that reliable benchmark use requires examining sample-level capability boundaries beyond leaderboard scores.
