Low Rank for Rank: Uncertainty-Aware Task-Specific LLM Ranking under Sparse Pairwise Comparisons

• URL: http://arxiv.org/abs/2605.29395v1
• Date: Thu, 28 May 2026 05:44:43 GMT
• Title: Low Rank for Rank: Uncertainty-Aware Task-Specific LLM Ranking under Sparse Pairwise Comparisons
• Abstract summary: Low-rank sharing improves sample efficiency over independent task-wise Bradley-Terry estimation.<n>Experiments on synthetic data and Arena show that low-rank sharing improves sample efficiency over independent task-wise Bradley-Terry estimation.
• Score: 16.73008463123115
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Pairwise human-preference platforms such as Chatbot Arena have become central to large language model (LLM) evaluation, yet reliable task-specific ranking remains challenging. Global leaderboards mask task heterogeneity, while ranking each fine-grained task independently is unstable under sparse, imbalanced comparisons. We propose a low-rank framework for task-specific LLM ranking from sparse pairwise comparisons, modeling the task-by-model ability matrix $Θ^\star \in \mathbb{R}^{d_t \times d_m}$ as low rank so that information is shared across related tasks while task-specific differences are preserved. We first develop a max-norm ($\ell_\infty$) accurate estimator for the latent scores, combining a convex initializer with alternating-minimization refinement, and prove task-wise top-$K$ recovery guarantees under sparse sampling. Our main contribution is an uncertainty quantification framework for task-specific ranking. We construct cross-fitted one-step debiased estimators for fixed score contrasts -- such as the task-specific ability gap between two models -- yielding asymptotically valid confidence intervals that attain the semiparametric efficiency bound. We then extend the inference to the high-dimensional ranking regime, where per-task ranks and top-$K$ membership are determined by many dependent score-gap hypotheses. Using Gaussian and multiplier-bootstrap calibration, we obtain simultaneous confidence sets for per-task ranks and valid top-$K$ membership tests across many tasks and models. Experiments on synthetic data and Chatbot Arena show that low-rank sharing improves sample efficiency over independent task-wise Bradley-Terry estimation and produces tighter, better-calibrated ranking certificates, with the largest gains in the sparse regime typical of real LLM benchmarks.

Related papers

• Task-CoEvolve: Efficient Harness Optimization via Adaptive Validation Task Selection [51.39973047614183]
  We present a novel approach to efficient LLM harness optimization through adaptive validation task selection.<n>$textbfTask-CoEvolve$, which co-evolves the validation tasks with the harness by addressing two challenges: selecting informative tasks and estimating full-set performance from partial evaluations.
  arXiv  Detail & Related papers  (2026-08-20T15:24:54Z)
• Rank Intervals for Leaderboards: A Hierarchical Framework for Model Evaluation [4.297070083645049]
  We introduce a hierarchical framework that constructs model rank intervals with statistical guarantees at both levels.<n>Experiments on simulated data and the TabArena and PromptEval benchmarks show that our method yields statistically valid and informative intervals.
  arXiv  Detail & Related papers  (2026-06-07T15:31:29Z)
• Multi-task Linear Regression without Eigenvalue Lower Bounds: Adaptivity, Robustness and Safety [2.538209532048867]
  We study the multi-task linear regression problem in the presence of contaminated tasks.<n>We introduce a relative balancedness condition, quantified by a balancedness constant, that compares each task's second moment with the average inlier geometry.<n>Our method achieves simultaneous adaptivity to task similarity, robustness to outliers, and safety outside favorable transfer regimes.
  arXiv  Detail & Related papers  (2026-05-16T19:06:54Z)
• Agent psychometrics: Task-level performance prediction in agentic coding benchmarks [24.348135523715815]
  We present a framework for predicting success or failure on individual tasks tailored to the agentic coding regime.<n>Our approach augments Item Response Theory (IRT) with rich features extracted from tasks, including issue statements, repository contexts, solutions, and test cases.
  arXiv  Detail & Related papers  (2026-04-01T07:59:59Z)
• Efficient Benchmarking of AI Agents [0.0]
  We study whether small task subsets can preserve agent rankings at substantially lower cost.<n>We find that absolute score prediction degrades under scaffold-driven distribution shift.<n>We propose an optimization-free protocol: evaluate new agents only on tasks with intermediate historical pass rates.
  arXiv  Detail & Related papers  (2026-03-24T22:17:11Z)
• ArenaRL: Scaling RL for Open-Ended Agents via Tournament-based Relative Ranking [84.07076200941474]
  ArenaRL is a reinforcement learning paradigm that shifts from pointwise scalar scoring to intra-group relative ranking.<n>We construct an intra-group adversarial arena and devise a tournament-based ranking scheme to obtain stable advantage signals.<n>Experiments show that ArenaRL substantially outperforms standard RL baselines.
  arXiv  Detail & Related papers  (2026-01-10T08:43:07Z)
• NDCG-Consistent Softmax Approximation with Accelerated Convergence [67.10365329542365]
  We propose novel loss formulations that align directly with ranking metrics.<n>We integrate the proposed RG losses with the highly efficient Alternating Least Squares (ALS) optimization method.<n> Empirical evaluations on real-world datasets demonstrate that our approach achieves comparable or superior ranking performance.
  arXiv  Detail & Related papers  (2025-06-11T06:59:17Z)
• From Continual Learning to SGD and Back: Better Rates for Continual Linear Models [50.11453013647086]
  We analyze the forgetting, i.e., loss on previously seen tasks, after $k$ iterations.<n>We develop novel last-iterate upper bounds in the realizable least squares setup.<n>We prove for the first time that randomization alone, with no task repetition, can prevent catastrophic in sufficiently long task sequences.
  arXiv  Detail & Related papers  (2025-04-06T18:39:45Z)
• AdaRank: Adaptive Rank Pruning for Enhanced Model Merging [23.649762835129167]
  Model merging has emerged as a promising approach for unifying independently fine-tuned models into an integrated framework.<n>We propose AdaRank, a novel model merging framework that adaptively selects the most beneficial singular directions of task vectors to merge multiple models.<n>AdaRank consistently achieves state-of-the-art performance with various backbones and number of tasks, reducing the performance gap between fine-tuned models to nearly 1%.
  arXiv  Detail & Related papers  (2025-03-28T06:49:06Z)
• Exposing and Addressing Cross-Task Inconsistency in Unified Vision-Language Models [80.23791222509644]
  Inconsistent AI models are considered brittle and untrustworthy by human users. We find that state-of-the-art vision-language models suffer from a surprisingly high degree of inconsistent behavior across tasks. We propose a rank correlation-based auxiliary training objective, computed over large automatically created cross-task contrast sets.
  arXiv  Detail & Related papers  (2023-03-28T16:57:12Z)
• New Tight Relaxations of Rank Minimization for Multi-Task Learning [161.23314844751556]
  We propose two novel multi-task learning formulations based on two regularization terms. We show that our methods can correctly recover the low-rank structure shared across tasks, and outperform related multi-task learning methods.
  arXiv  Detail & Related papers  (2021-12-09T07:29:57Z)

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.