RewardHarness: Self-Evolving Agentic Post-Training

• URL: http://arxiv.org/abs/2605.08703v1
• Date: Sat, 09 May 2026 05:32:48 GMT
• Title: RewardHarness: Self-Evolving Agentic Post-Training
• Abstract summary: RewardHarness is a self-evolving agentic reward framework for instruction-guided image editing.<n>It aligns with human preferences by iteratively evolving a library of tools and skills from as few as 100 preference demonstrations.<n>It achieves 47.4% average accuracy on image-editing evaluation benchmarks, surpassing GPT-5 by 5.3 points.
• Score: 51.16917192137169
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Evaluating instruction-guided image edits requires rewards that reflect subtle human preferences, yet current reward models typically depend on large-scale preference annotation and additional model training. This creates a data-efficiency gap: humans can often infer the target evaluation criteria from only a few examples, while models are usually trained on hundreds of thousands of comparisons. We present RewardHarness, a self-evolving agentic reward framework that reframes reward modeling as context evolution rather than weight optimization. Instead of learning from large-scale annotations, RewardHarness aligns with human preferences by iteratively evolving a library of tools and skills from as few as 100 preference demonstrations. Given a source image, candidate edited images, and an editing instruction, an Orchestrator selects the most relevant subset of tools and skills from the maintained library, and a frozen Sub-Agent uses them to construct a reasoning chain that produces a preference judgment. By comparing predicted judgments with ground-truth preferences and analyzing successes and failures in the reasoning process, the Orchestrator automatically refines its library of tools and skills without additional human annotation. Using only 0.05% of the EditReward preference data, RewardHarness achieves 47.4% average accuracy on image-editing evaluation benchmarks, surpassing GPT-5 by 5.3 points. When used as a reward signal for GRPO fine-tuning, RL-tuned models achieve 3.52 on ImgEdit-Bench. Project page: https://rewardharness.com.

Related papers

• ReasonEdit: Towards Interpretable Image Editing Evaluation via Reinforcement Learning [86.61218827780675]
  We present ReasonEdit, an evaluation tool for text-guided image editing.<n>It is trained using reward signals derived from RE-Reward and the Group Relative Policy Optimization (GRPO) algorithm.<n>It is capable of generating high-quality interpretable evaluation text, enabling more transparent and trustworthy assessment for image editing.
  arXiv  Detail & Related papers  (2026-05-08T09:23:26Z)
• RationalRewards: Reasoning Rewards Scale Visual Generation Both Training and Test Time [51.256131853751754]
  We show that teaching reward models to produce explicit, multi-dimensional critiques transforms them from passive evaluators into active optimization tools.<n>We introduce Preference-Anchored Rationalization (PARROT), a principled framework that recovers high-quality rationales from readily available preference data.<n>The resulting model, RationalRewards (8B), achieves state-of-the-art preference prediction among open-source reward models.
  arXiv  Detail & Related papers  (2026-04-13T15:38:09Z)
• One Model to Critique Them All: Rewarding Agentic Tool-Use via Efficient Reasoning [54.580646706013965]
  Reward models (RMs) play a critical role in aligning large language models with human preferences.<n>We introduce ToolRM, a family of lightweight generative RMs tailored for general tool-use scenarios.<n>To build these models, we propose a novel pipeline that constructs pairwise preference data using rule-based scoring and multidimensional sampling.
  arXiv  Detail & Related papers  (2025-10-30T06:08:27Z)
• Listener-Rewarded Thinking in VLMs for Image Preferences [38.07052490646366]
  We introduce a listener-augmented GRPO framework to train visual reward models.<n>Our listener-shaped reward scheme achieves best accuracy on the ImageReward benchmark.<n>These results demonstrate that listener-based rewards provide a scalable, data-efficient path to aligning vision-language models with nuanced human preferences.
  arXiv  Detail & Related papers  (2025-06-28T09:53:17Z)
• Capturing Individual Human Preferences with Reward Features [47.43999785878563]
  We show that individual preferences can be captured as a linear combination of a set of general reward features.<n>We show how to learn such features and subsequently use them to quickly adapt the reward model to a specific individual.<n>We present experiments with large language models comparing the proposed architecture with a non-adaptive reward model and also adaptive counterparts.
  arXiv  Detail & Related papers  (2025-03-21T17:39:33Z)
• Self-Taught Evaluators [77.92610887220594]
  We present an approach that aims to im-proves without human annotations, using synthetic training data only. Our Self-Taught Evaluator can improve a strong LLM from 75.4 to 88.3 on RewardBench.
  arXiv  Detail & Related papers  (2024-08-05T17:57:02Z)
• RewardBench: Evaluating Reward Models for Language Modeling [100.28366840977966]
  We present RewardBench, a benchmark dataset and code-base for evaluation of reward models. The dataset is a collection of prompt-chosen-rejected trios spanning chat, reasoning, and safety. On the RewardBench leaderboard, we evaluate reward models trained with a variety of methods.
  arXiv  Detail & Related papers  (2024-03-20T17:49:54Z)

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.