Aligning Language Model Benchmarks with Pairwise Preferences

• URL: http://arxiv.org/abs/2602.02898v1
• Date: Mon, 02 Feb 2026 23:11:09 GMT
• Title: Aligning Language Model Benchmarks with Pairwise Preferences
• Authors: Marco Gutierrez, Xinyi Leng, Hannah Cyberey, Jonathan Richard Schwarz, Ahmed Alaa, Thomas Hartvigsen,
• Abstract summary: We introduce benchmark alignment, where we use limited amounts of information about model performance to automatically update offline benchmarks.<n>We then propose BenchAlign, which learns preference-aligned weight-ings for benchmark questions.<n>Our experiments show that our aligned benchmarks can accurately rank unseen models according to models of human preferences, even across different sizes.
• Score: 15.427340427081843
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Language model benchmarks are pervasive and computationally-efficient proxies for real-world performance. However, many recent works find that benchmarks often fail to predict real utility. Towards bridging this gap, we introduce benchmark alignment, where we use limited amounts of information about model performance to automatically update offline benchmarks, aiming to produce new static benchmarks that predict model pairwise preferences in given test settings. We then propose BenchAlign, the first solution to this problem, which learns preference-aligned weight- ings for benchmark questions using the question-level performance of language models alongside ranked pairs of models that could be collected during deployment, producing new benchmarks that rank previously unseen models according to these preferences. Our experiments show that our aligned benchmarks can accurately rank unseen models according to models of human preferences, even across different sizes, while remaining interpretable. Overall, our work provides insights into the limits of aligning benchmarks with practical human preferences, which stands to accelerate model development towards real utility.

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)
• 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)
• Scales++: Compute Efficient Evaluation Subset Selection with Cognitive Scales Embeddings [23.9553588103042]
  We propose a item-centric approach to benchmark subset selection, arguing that selection should be based on the intrinsic properties of the task items themselves.<n>We show Scales++ reduces the upfront selection cost by over 18x while achieving competitive predictive fidelity.<n>We demonstrate that this item-centric approach enables more efficient model evaluation without significant fidelity degradation.
  arXiv  Detail & Related papers  (2025-10-30T11:28:58Z)
• RewardBench 2: Advancing Reward Model Evaluation [71.65938693914153]
  Reward models are used throughout the post-training of language models to capture nuanced signals from preference data.<n>The community has begun establishing best practices for evaluating reward models.<n>This paper introduces RewardBench 2, a new multi-skill reward modeling benchmark.
  arXiv  Detail & Related papers  (2025-06-02T17:54:04Z)
• LiveXiv -- A Multi-Modal Live Benchmark Based on Arxiv Papers Content [62.816876067499415]
  We propose LiveXiv: a scalable evolving live benchmark based on scientific ArXiv papers.<n>LiveXiv accesses domain-specific manuscripts at any given timestamp and proposes to automatically generate visual question-answer pairs.<n>We benchmark multiple open and proprietary Large Multi-modal Models (LMMs) on the first version of our benchmark, showing its challenging nature and exposing the models true abilities.
  arXiv  Detail & Related papers  (2024-10-14T17:51:23Z)
• Margin Matching Preference Optimization: Enhanced Model Alignment with Granular Feedback [64.67540769692074]
  Large language models (LLMs) fine-tuned with alignment techniques, such as reinforcement learning from human feedback, have been instrumental in developing some of the most capable AI systems to date.<n>We introduce an approach called Margin Matching Preference Optimization (MMPO), which incorporates relative quality margins into optimization, leading to improved LLM policies and reward models.<n>Experiments with both human and AI feedback data demonstrate that MMPO consistently outperforms baseline methods, often by a substantial margin, on popular benchmarks including MT-bench and RewardBench.
  arXiv  Detail & Related papers  (2024-10-04T04:56:11Z)
• Beyond Bradley-Terry Models: A General Preference Model for Language Model Alignment [51.14207112118503]
  We introduce preference embedding, an approach that embeds responses into a latent space to capture preferences efficiently.<n>We also propose preference score-based General Preference Optimization (GPO), which generalizes reward-based reinforcement learning from human feedback.<n>Our method may enhance the alignment of foundation models with nuanced human values.
  arXiv  Detail & Related papers  (2024-10-03T04:22:55Z)
• Anchor Points: Benchmarking Models with Much Fewer Examples [88.02417913161356]
  In six popular language classification benchmarks, model confidence in the correct class on many pairs of points is strongly correlated across models. We propose Anchor Point Selection, a technique to select small subsets of datasets that capture model behavior across the entire dataset. Just several anchor points can be used to estimate model per-class predictions on all other points in a dataset with low mean absolute error.
  arXiv  Detail & Related papers  (2023-09-14T17:45:51Z)

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.