Duel-Evolve: Reward-Free Test-Time Scaling via LLM Self-Preferences

• URL: http://arxiv.org/abs/2602.21585v1
• Date: Wed, 25 Feb 2026 05:16:11 GMT
• Title: Duel-Evolve: Reward-Free Test-Time Scaling via LLM Self-Preferences
• Abstract summary: We introduce Duel-Evolve, an evolutionary algorithm that replaces external scalar rewards with pairwise preferences elicited from the same LLM used to generate candidates.<n>Duel-Evolve aggregates noisy candidate comparisons via a Bayesian Bradley-Terry model, yielding uncertainty-aware estimates of candidate quality.<n>We evaluate Duel-Evolve on MathBench, where it achieves 20 percentage points higher accuracy over existing methods and baselines, and on LiveCodeBench, where it improves over comparable iterative methods by over 12 percentage points.
• Score: 7.715856473694668
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Many applications seek to optimize LLM outputs at test time by iteratively proposing, scoring, and refining candidates over a discrete output space. Existing methods use a calibrated scalar evaluator for the target objective to guide search, but for many tasks such scores are unavailable, too sparse, or unreliable. Pairwise comparisons, by contrast, are often easier to elicit, still provide useful signal on improvement directions, and can be obtained from the LLM itself without external supervision. Building on this observation, we introduce Duel-Evolve, an evolutionary optimization algorithm that replaces external scalar rewards with pairwise preferences elicited from the same LLM used to generate candidates. Duel-Evolve aggregates these noisy candidate comparisons via a Bayesian Bradley-Terry model, yielding uncertainty-aware estimates of candidate quality. These quality estimates guide allocation of the comparison budget toward plausible optima using Double Thompson Sampling, as well as selection of high-quality parents to generate improved candidates. We evaluate Duel-Evolve on MathBench, where it achieves 20 percentage points higher accuracy over existing methods and baselines, and on LiveCodeBench, where it improves over comparable iterative methods by over 12 percentage points. Notably, the method requires no reward model, no ground-truth labels during search, and no hand-crafted scoring function. Results show that pairwise self-preferences provide strong optimization signal for test-time improvement over large, discrete output spaces.

Related papers

• Refining Over Resampling: Test-Time Self-Correction for LLM Reasoning [3.8191765711200065]
  We propose a verifier-free breadth--depth refinement framework that uses test-time compute to both explore and improve candidate solutions.<n>Our method consistently improves over greedy decoding, majority voting, verifier-based best-of-$N$, beam search, and look decoding across multiple open-weight models.
  arXiv  Detail & Related papers  (2026-08-06T06:38:37Z)
• LLM-as-a-Verifier: A General-Purpose Verification Framework [74.40111651545979]
  We introduce LLM-as-a-Verifier, a general-purpose verification framework.<n>It provides fine-grained feedback for agentic tasks without requiring additional training.<n>It achieves state-of-the-art performance on several benchmarks.
  arXiv  Detail & Related papers  (2026-07-06T17:59:35Z)
• $S^3$: Stratified Scaling Search for Test-Time in Diffusion Language Models [10.157422365382933]
  Test-time scaling investigates whether a fixed diffusion language model (DLM) can generate better outputs when given more inference compute.<n>We propose $S3$ (Stratified Scaling Search), a verifier-guided search method that improves generation by reallocating compute during the denoising process.<n>Experiments with LLaDA-8B-Instruct on MATH-500, GSM8K, ARC-Challenge, and TruthfulQA demonstrate that $S3$ consistently improves performance across benchmarks.
  arXiv  Detail & Related papers  (2026-04-07T00:51:06Z)
• $\ abla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Latent Space [71.23672814629448]
  $nabla$-Reasoner is an iterative generation framework that integrates differentiable optimization over token logits into the decoding loop.<n>$nabla$-Reasoner achieves over 20% accuracy improvement on a challenging mathematical reasoning benchmark.
  arXiv  Detail & Related papers  (2026-03-05T08:42:54Z)
• GenSelect: A Generative Approach to Best-of-N [46.74684914760564]
  We introduce GenSelect, where the LLM uses long reasoning to select the best solution among N candidates.<n>For math reasoning, we demonstrate that reasoning models, such as QwQ and DeepSeek-R1-0528, excel at GenSelect.
  arXiv  Detail & Related papers  (2025-07-23T15:22:51Z)
• PMPO: Probabilistic Metric Prompt Optimization for Small and Large Language Models [1.6816171955882597]
  PMPO locates low quality prompt segments via a masking based analysis and iteratively rewrites them to propose improved variants.<n>It selects among variants by minimizing loss in a single forward pass, eliminating output sampling and human or judge based scoring for selection.<n>Across model sizes and datasets, PMPO outperforms prior prompts: it achieves the highest average accuracy on BBH, performs strongly on GSM8K and AQUA RAT, and raises AlpacaEval 2.0 win rates by over 19 points.
  arXiv  Detail & Related papers  (2025-05-22T06:59:10Z)
• Sample, Don't Search: Rethinking Test-Time Alignment for Language Models [55.2480439325792]
  We introduce QAlign, a new test-time alignment approach.<n>As we scale test-time compute, QAlign converges to sampling from the optimal aligned distribution for each individual prompt.<n>By adopting recent advances in Markov chain Monte Carlo for text generation, our method enables better-aligned outputs without modifying the underlying model or even requiring logit access.
  arXiv  Detail & Related papers  (2025-04-04T00:41:40Z)
• Scalable Best-of-N Selection for Large Language Models via Self-Certainty [75.1351701045874]
  Best-of-N selection is a key technique for improving the reasoning performance of Large Language Models (LLMs)<n>We propose self-certainty, a novel and efficient metric that leverages the inherent probability distribution of LLM outputs to estimate response quality without requiring external reward models.<n>Our findings establish self-certainty as a practical and efficient way for improving LLM reasoning capabilities.
  arXiv  Detail & Related papers  (2025-02-25T19:08:07Z)
• MM-RLHF: The Next Step Forward in Multimodal LLM Alignment [59.536850459059856]
  We introduce MM-RLHF, a dataset containing $mathbf120k$ fine-grained, human-annotated preference comparison pairs.<n>We propose several key innovations to improve the quality of reward models and the efficiency of alignment algorithms.<n>Our approach is rigorously evaluated across $mathbf10$ distinct dimensions and $mathbf27$ benchmarks.
  arXiv  Detail & Related papers  (2025-02-14T18:59:51Z)
• Reward-Augmented Data Enhances Direct Preference Alignment of LLMs [63.32585910975191]
  We introduce reward-conditioned Large Language Models (LLMs) that learn from the entire spectrum of response quality within the dataset.<n>We show that our approach consistently boosts DPO by a considerable margin.<n>Our method not only maximizes the utility of preference data but also mitigates the issue of unlearning, demonstrating its broad effectiveness beyond mere data expansion.
  arXiv  Detail & Related papers  (2024-10-10T16:01:51Z)
• Self-Exploring Language Models: Active Preference Elicitation for Online Alignment [88.56809269990625]
  We propose a bilevel objective optimistically biased towards potentially high-reward responses to actively explore out-of-distribution regions. Our experimental results demonstrate that when fine-tuned on Zephyr-7B-SFT and Llama-3-8B-Instruct models, Self-Exploring Language Models (SELM) significantly boosts the performance on instruction-following benchmarks.
  arXiv  Detail & Related papers  (2024-05-29T17:59:07Z)
• PRD: Peer Rank and Discussion Improve Large Language Model based Evaluations [10.709365940160685]
  Modern large language models (LLMs) are hard to evaluate and compare automatically.<n>We propose a peer rank (PR) algorithm that takes into account each peer LLM's pairwise preferences of all answer pairs.<n>We find that our approaches achieve higher accuracy and align better with human judgments.
  arXiv  Detail & Related papers  (2023-07-06T04:05:44Z)
• Let's Sample Step by Step: Adaptive-Consistency for Efficient Reasoning and Coding with LLMs [60.58434523646137]
  A popular approach for improving the correctness of output from large language models (LLMs) is Self-Consistency. We introduce Adaptive-Consistency, a cost-efficient, model-agnostic technique that dynamically adjusts the number of samples per question. Our experiments show that Adaptive-Consistency reduces sample budget by up to 7.9 times with an average accuracy drop of less than 0.1%.
  arXiv  Detail & Related papers  (2023-05-19T17:49:25Z)
• Local policy search with Bayesian optimization [73.0364959221845]
  Reinforcement learning aims to find an optimal policy by interaction with an environment. Policy gradients for local search are often obtained from random perturbations. We develop an algorithm utilizing a probabilistic model of the objective function and its gradient.
  arXiv  Detail & Related papers  (2021-06-22T16:07:02Z)

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.