paper-with-me

홈 › Papers

Visual-ERM: Reward Modeling for Visual Equivalence

2026-03-13 · Ziyu Liu, Shengyuan Ding, Xinyu Fang, Xuanlang Dai, Penghui Yang, Jianze Liang, Jiaqi Wang, Kai Chen, Dahua Lin, Yuhang Zang arxiv

Vision-to-code tasks require models to reconstruct structured visual inputs, such as charts, tables, and SVGs, into executable or structured representations with high visual fidelity. While recent Large Vision Language Models (LVLMs) achieve strong results via supervised fine-tuning, reinforcement learning remains challenging due to misaligned reward signals. Existing rewards either rely on textual rules or coarse visual embedding similarity, both of which fail to capture fine-grained visual discrepancies and are vulnerable to reward hacking. We propose Visual Equivalence Reward Model (Visual-ERM), a multimodal generative reward model that provides fine-grained, interpretable, and task-agnostic feedback to evaluate vision-to-code quality directly in the rendered visual space. Integrated into RL, Visual-ERM improves Qwen3-VL-8B-Instruct by +8.4 on chart-to-code and yields consistent gains on table and SVG parsing (+2.7, +4.1 on average), and further strengthens test-time scaling via reflection and revision. We also introduce VisualCritic-RewardBench (VC-RewardBench), a benchmark for judging fine-grained image-to-image discrepancies on structured visual data, where Visual-ERM at 8B decisively outperforms Qwen3-VL-235B-Instruct and approaches leading closed-source models. Our results suggest that fine-grained visual reward supervision is both necessary and sufficient for vision-to-code RL, regardless of task specificity.

📄 PDF Abstract BibTeX arXiv:2603.13224

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

ViGoR: Improving Visual Grounding of Large Vision Language Models with Fine-Grained Reward Modeling

2024-02-09 · Siming Yan, Min Bai, Weifeng Chen, Xiong Zhou 외

By combining natural language understanding, generation capabilities, and breadth of knowledge of large language models with image perception, recent large vision language models (LVLMs) have shown unprecedented visual r…

HallucinationNatural Language UnderstandingVisual GroundingVisual Reasoning

ARM-Thinker: Reinforcing Multimodal Generative Reward Models with Agentic Tool Use and Visual Reasoning

2025-12-04 · Shengyuan Ding, Xinyu Fang, Ziyu Liu, Yuhang Zang 외 arxiv

Reward models are critical for aligning vision-language systems with human preferences, yet current approaches suffer from hallucination, weak visual grounding, and an inability to use tools for verification, limiting th…

Reinforcement LearningInstruction FollowingMultimodal ReasoningLogical Reasoning

Implicit Affordance Acquisition via Causal Action-Effect Modeling in the Video Domain

2023-12-18 · Hsiu-Yu Yang, Carina Silberer

Affordance knowledge is a fundamental aspect of commonsense knowledge. Recent findings indicate that world knowledge emerges through large-scale self-supervised pretraining, motivating our exploration of acquiring afford…

World Knowledge

Unified Personalized Reward Model for Vision Generation

2026-02-02 · Yibin Wang, Yuhang Zang, Feng Han, Jiazi Bu 외 arxiv

Recent advancements in multimodal reward models (RMs) have significantly propelled the development of visual generation. Existing frameworks typically adopt Bradley-Terry-style preference modeling or leverage generative …

Reinforcement Learning

GENNI: Visualising the Geometry of Equivalences for Neural Network Identifiability

2020-11-14 · Daniel Lengyel, Janith Petangoda, Isak Falk, Kate Highnam 외

We propose an efficient algorithm to visualise symmetries in neural networks. Typically, models are defined with respect to a parameter space, where non-equal parameters can produce the same input-output map. Our propose…