Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.23255

Sample, Mix, or Verify? Allocating Test-Time Compute under Imperfect Verifiers

Srinivasan Iyengar
View original source ↗Read source document ↗Export metadata ↓

Abstract

A common way to make language models and agents more accurate is to spend more compute at inference time: generate many candidate solutions and submit the one a verifier prefers. Success then requires that some candidate is correct and that the verifier picks it. The first grows predictably with the number of samples; the second does not, because real verifiers (unit tests, majority votes, LLM judges, reward models) accept wrong answers, reject right ones, and make correlated mistakes. With a fixed budget, a practitioner must decide how many samples to draw, which models to draw them from, and how much to spend on verification, and today these choices are made by rules of thumb. We introduce SPLIT, which makes them from a small labelled pilot. It rests on an exact formula for the probability of submitting a correct answer under accept/reject and top-score selection that accounts for varying task difficulty, blind spots shared across models, and correlated repeated verifier calls. SPLIT fits the formula by empirical Bayes and uses it to plan budgets on unseen tasks. Across seven benchmarks spanning math, science QA, code and software-engineering agents, 26 models, 175 agent systems and six kinds of real verifiers, all from public releases, SPLIT lands within 0.02-0.8 percentage points of the best allocation in hindsight, while every fixed rule we test loses at least 2.5 points in some setting. Regret depends only on how well SPLIT ranks configurations, and it stays near-optimal under compute, price and latency costs and online. The formula also explains when asking more LLM judges lowers precision. As a forecaster far beyond the pilot, SPLIT beats power-law fitting when the verifier caps success, and we report where it loses.

Explore related subjects

cs.LGcs.CL

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Srinivasan Iyengar. 2026-10-16. Sample, Mix, or Verify? Allocating Test-Time Compute under Imperfect Verifiers. https://arxiv.org/abs/2610.23255

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Is the Policy Gradient a Gradient?

The policy gradient theorem describes the gradient of the expected discounted return with respect to an agent's policy parameters. However, many policy gradient methods omit the outer discount factor, yielding an update that need not equal the gradient of the discounted objective. Is the resulting update the gradient of another objective? This has been an open question for several years, and this lack of theoretical clarity has led to an abundance of misstatements in the literature. We answer this question for the resulting episode-summed, on-policy update by proving that it need not equal the gradient of any function. Further, we construct a counterexample in which exact expected updates drive the policy toward a deterministic limit that is globally pessimal with respect to both the discounted and undiscounted objectives, while the parameter diverges. We motivate this work by surveying the literature and showing that there remains a widespread misunderstanding regarding discounted policy gradient methods, with errors present even in highly-cited papers published at top conferences.

cs.LG↗
arXiv2026

Harnessing Trust in Directed Graphs: Redefining Robustness of Graph Learning

Existing research on robust Graph Neural Networks (GNNs) focuses predominantly on undirected graphs, overlooking the trustworthiness inherent in the link directions of directed graphs. In this work, we analyze the limitations of current approaches from both the attack and defense perspectives and redefine the robustness of GNNs in directed graphs. From the attack side, we introduce a more realistic directed graph attack setting that addresses the shortcomings of existing attacks. From the defense side, we propose a simple yet effective message-passing framework that serves as a plug-in layer to strengthen GNN robustness while avoiding a false sense of robustness. Our findings show that the trust signals encoded in directed graphs can be harnessed to substantially improve the robustness of GNNs. When combined with existing defense strategies, the framework achieves state-of-the-art robust performance alongside strong clean accuracy against both transfer and adaptive attacks. These results point to a novel and promising direction for strengthening the robustness of graph deep learning.

cs.LG↗
arXiv2026

Two Minds Better Than One: Collaborative Reward Modeling for LLM Alignment

Reward models (RMs) are central to aligning large language models (LLMs) with human values, yet noisy preference feedback can induce reward misgeneralization through spurious correlations and overfitting. We characterize the training dynamics of preference pairs and uncover a distinctive optimization signature of noise: persistently high losses, sharp temporal fluctuations, and irregular gradient updates. Our analyses further show that excluding such pairs improves downstream alignment. In practice, however, reliable preferences are not known a priori, and self-filtering based on a model's own training signals can amplify confirmation bias. To address this challenge, we introduce Collaborative Reward Modeling (CRM), an online framework for identifying reliable preferences. CRM couples cross-model peer review with curriculum learning: two RMs evaluate each other's sample selections, while curriculum keeps their learning progress synchronized. Across extensive experiments, CRM improves RM generalization by up to 9.94 points on RewardBench under 40% preference noise. CRM further generalizes to implicit-reward alignment, offering a robust and broadly applicable approach to learning from imperfect feedback.

cs.LG↗