Benchmark Profiling: Mechanistic Diagnosis of LLM Benchmarks

• URL: http://arxiv.org/abs/2510.01232v1
• Date: Tue, 23 Sep 2025 15:32:47 GMT
• Title: Benchmark Profiling: Mechanistic Diagnosis of LLM Benchmarks
• Authors: Dongjun Kim, Gyuho Shim, Yongchan Chun, Minhyuk Kim, Chanjun Park, Heuiseok Lim,
• Abstract summary: Benchmark Profiling decomposes benchmark performance into ten cognitively grounded abilities.<n>It explains why performance gains do not always translate into user-perceived competence.
• Score: 34.09939383415074
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Language Models are commonly judged by their scores on standard benchmarks, yet such scores often overstate real capability since they mask the mix of skills a task actually demands. For example, ARC is assumed to test reasoning, while HellaSwag is designed to evaluate commonsense. However, we lack a systematic way to verify if these benchmarks actually measure these labels. We introduce Benchmark Profiling, a diagnostic framework that decomposes benchmark performance into ten cognitively grounded abilities. The method combines gradient-based importance scoring with targeted parameter ablation to compute an Ability Impact Score (AIS) that quantifies how much each ability contributes to a model's success on a given benchmark. Profiling three instruction-tuned models across ten widely used benchmarks yields four key findings: (i) most benchmarks draw on several abilities rather than one, (ii) datasets with similar labels rely on distinct ability mixtures, (iii) code-generation benchmarks reward broad, multi-skill improvement and thus show only modest gains from narrow domain-specific fine-tuning, and (iv) abilities irrelevant to the task could negatively affect performance. Benchmark Profiling therefore explains why performance gains do not always translate into user-perceived competence and offers a transparent tool for benchmark audit and model interpretability.

Related papers

• IF-RewardBench: Benchmarking Judge Models for Instruction-Following Evaluation [85.56193980646981]
  We propose IF-RewardBench, a comprehensive meta-evaluation benchmark for instruction-following.<n>For each instruction, we construct a preference graph containing all pairwise preferences among multiple responses.<n>Experiments on IF-RewardBench reveal significant deficiencies in current judge models.
  arXiv  Detail & Related papers  (2026-03-05T02:21:17Z)
• When AI Benchmarks Plateau: A Systematic Study of Benchmark Saturation [80.66788281323414]
  We analyze benchmark saturation across 60 Large Language Model (LLM) benchmarks selected from technical reports by major model developers.<n>Our analysis reveals that nearly half of the benchmarks exhibit saturation, with rates increasing as benchmarks age.<n>Expert-curated benchmarks resist saturation better than crowdsourced ones.
  arXiv  Detail & Related papers  (2026-02-18T16:51:37Z)
• Benchmark^2: Systematic Evaluation of LLM Benchmarks [66.2731798872668]
  We propose Benchmark2, a comprehensive framework comprising three complementary metrics.<n>We conduct experiments across 15 benchmarks spanning mathematics, reasoning, and knowledge domains.<n>Our analysis reveals significant quality variations among existing benchmarks and demonstrates that selective benchmark construction can achieve comparable evaluation performance.
  arXiv  Detail & Related papers  (2026-01-07T14:59:03Z)
• Uncovering Competency Gaps in Large Language Models and Their Benchmarks [11.572508874955659]
  We propose a new method that uses sparse autoencoders (SAEs) to automatically uncover both types of gaps.<n>We found that models consistently underperformed on concepts that stand in contrast to sycophantic behaviors.<n>Our method offers a representation-grounded approach to evaluation, enabling concept-level decomposition of benchmark scores.
  arXiv  Detail & Related papers  (2025-12-06T17:39:47Z)
• Fantastic Bugs and Where to Find Them in AI Benchmarks [28.604919035475188]
  We introduce a framework for systematic benchmark revision that leverages statistical analysis of response patterns to flag potentially invalid questions.<n>Our approach builds on a core assumption commonly used in AI evaluations that the mean score sufficiently summarizes model performance.<n>Across nine widely used benchmarks, our method guides expert review to identify problematic questions with up to 84% precision.
  arXiv  Detail & Related papers  (2025-11-20T22:49:21Z)
• Benchmark Designers Should "Train on the Test Set" to Expose Exploitable Non-Visual Shortcuts [49.99400612296149]
  We find that models can ace many benchmarks without strong visual understanding.<n>This is especially problematic for vision-centric benchmarks that are meant to require visual inputs.<n>We adopt a diagnostic principle for benchmark design: if a benchmark can be gamed, it will be gamed.
  arXiv  Detail & Related papers  (2025-11-06T18:43:21Z)
• Fluid Language Model Benchmarking [126.92394365620525]
  We introduce Fluid Benchmarking, a new evaluation approach that advances LM benchmarking across multiple dimensions.<n>Inspired by psychometrics, Fluid Benchmarking is based on the insight that the relative value of benchmark items depends on an LM's capability level.<n>We examine four dimensions -- efficiency, validity, variance, and saturation -- and find that Fluid Benchmarking achieves superior performance in all of them.
  arXiv  Detail & Related papers  (2025-09-14T05:49:42Z)
• Beyond the Singular: The Essential Role of Multiple Generations in Effective Benchmark Evaluation and Analysis [10.133537818749291]
  Large language models (LLMs) have demonstrated significant utilities in real-world applications.<n> Benchmark evaluations are crucial for assessing the capabilities of LLMs.
  arXiv  Detail & Related papers  (2025-02-13T03:43:33Z)
• Do Large Language Model Benchmarks Test Reliability? [66.1783478365998]
  We investigate how well current benchmarks quantify model reliability.<n>Motivated by this gap in the evaluation of reliability, we propose the concept of so-called platinum benchmarks.<n>We evaluate a wide range of models on these platinum benchmarks and find that, indeed, frontier LLMs still exhibit failures on simple tasks.
  arXiv  Detail & Related papers  (2025-02-05T18:58:19Z)
• Improving LLM Leaderboards with Psychometrical Methodology [0.0]
  The rapid development of large language models (LLMs) has necessitated the creation of benchmarks to evaluate their performance.<n>These benchmarks resemble human tests and surveys, as they consist of questions designed to measure emergent properties in the cognitive behavior of these systems.<n>However, unlike the well-defined traits and abilities studied in social sciences, the properties measured by these benchmarks are often vaguer and less rigorously defined.
  arXiv  Detail & Related papers  (2025-01-27T21:21:46Z)
• Do These LLM Benchmarks Agree? Fixing Benchmark Evaluation with BenchBench [15.565644819269803]
  We show how some overlooked methodological choices can significantly influence Benchmark Agreement Testing (BAT) results. We introduce BenchBench, a python package for BAT, and release the BenchBench-leaderboard, a meta-benchmark designed to evaluate benchmarks using their peers.
  arXiv  Detail & Related papers  (2024-07-18T17:00:23Z)
• MR-Ben: A Meta-Reasoning Benchmark for Evaluating System-2 Thinking in LLMs [55.20845457594977]
  Large language models (LLMs) have shown increasing capability in problem-solving and decision-making.<n>We present a process-based benchmark MR-Ben that demands a meta-reasoning skill.<n>Our meta-reasoning paradigm is especially suited for system-2 slow thinking.
  arXiv  Detail & Related papers  (2024-06-20T03:50:23Z)
• ECBD: Evidence-Centered Benchmark Design for NLP [95.50252564938417]
  We propose Evidence-Centered Benchmark Design (ECBD), a framework which formalizes the benchmark design process into five modules. Each module requires benchmark designers to describe, justify, and support benchmark design choices. Our analysis reveals common trends in benchmark design and documentation that could threaten the validity of benchmarks' measurements.
  arXiv  Detail & Related papers  (2024-06-13T00:59:55Z)
• Don't Make Your LLM an Evaluation Benchmark Cheater [142.24553056600627]
  Large language models(LLMs) have greatly advanced the frontiers of artificial intelligence, attaining remarkable improvement in model capacity. To assess the model performance, a typical approach is to construct evaluation benchmarks for measuring the ability level of LLMs. We discuss the potential risk and impact of inappropriately using evaluation benchmarks and misleadingly interpreting the evaluation results.
  arXiv  Detail & Related papers  (2023-11-03T14:59:54Z)
• The Benchmark Lottery [114.43978017484893]
  "A benchmark lottery" describes the overall fragility of the machine learning benchmarking process. We show that the relative performance of algorithms may be altered significantly simply by choosing different benchmark tasks.
  arXiv  Detail & Related papers  (2021-07-14T21:08:30Z)
• Interpretable Meta-Measure for Model Performance [4.91155110560629]
  We introduce a new meta-score assessment named Elo-based Predictive Power (EPP) EPP is built on top of other performance measures and allows for interpretable comparisons of models. We prove the mathematical properties of EPP and support them with empirical results of a large scale benchmark on 30 classification data sets and a real-world benchmark for visual data.
  arXiv  Detail & Related papers  (2020-06-02T14:10:13Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.