Training Vision-Language Process Reward Models for Test-Time Scaling in Multimodal Reasoning: Key Insights and Lessons Learned

• URL: http://arxiv.org/abs/2509.23250v3
• Date: Tue, 07 Oct 2025 02:40:16 GMT
• Title: Training Vision-Language Process Reward Models for Test-Time Scaling in Multimodal Reasoning: Key Insights and Lessons Learned
• Authors: Brandon Ong, Tej Deep Pala, Vernon Toh, William Chandra Tjhi, Soujanya Poria,
• Abstract summary: Process Reward Models (PRMs) improve the reliability of reasoning in large language models.<n>Existing Vision-Language PRMs rely on Monte Carlo Tree Search (MCTS) for data construction.<n>We introduce a hybrid data framework that combines MCTS with judgments from a strong VLM, producing more accurate step-level labels.
• Score: 29.44294456857936
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Process Reward Models (PRMs) provide step-level supervision that improves the reliability of reasoning in large language models. While PRMs have been extensively studied in text-based domains, their extension to Vision Language Models (VLMs) remains limited. Existing Vision-Language PRMs (VL-PRMs) rely on Monte Carlo Tree Search (MCTS) for data construction, which can often produce noisy supervision signals and limit generalization across tasks. In this work, we aim to elucidate the design space of VL-PRMs by exploring diverse strategies for dataset construction, training, and test-time scaling. First, we introduce a hybrid data synthesis framework that combines MCTS with judgments from a strong VLM, producing more accurate step-level labels. Second, we propose perception-focused supervision, enabling our PRM to explicitly detect errors at the visual grounding stage of reasoning. Third, we systematically evaluate multiple test-time scaling strategies, showing that our PRMs can reliably guide VLMs toward more accurate solutions. Our experiments covering five diverse multimodal benchmarks (MMMU, PuzzleVQA, AlgoPuzzleVQA, MathVista, and MathVision) reveal several key insights: (i) VL-PRMs when used as Outcome Reward Models (ORMs) during test-time scaling (TTS) can outperform VL-PRM guided process step selection, (ii) smaller VL-PRMs can match or even surpass larger ones in detecting process errors, (iii) VL-PRMs uncover latent reasoning abilities in stronger VLM backbones, (iv) perception-level supervision leads to significant gains in test-time scaling, and (v) TTS performance of different policies improve on advanced math reasoning datasets despite not training VL-PRMs on such datasets. We hope our work will motivate further research and support the advancement of VLMs.

Related papers

• Co-Training Vision Language Models for Remote Sensing Multi-task Learning [68.15604397741753]
  Vision language models (VLMs) have achieved promising results in RS image understanding, grounding, and ultra-high-resolution (UHR) image reasoning.<n>We present RSCoVLM, a simple yet flexible VLM baseline for RS MTL.<n>We propose a unified dynamic-resolution strategy to address the diverse image scales inherent in RS imagery.
  arXiv  Detail & Related papers  (2025-11-26T10:55:07Z)
• Boosting Multi-modal Keyphrase Prediction with Dynamic Chain-of-Thought in Vision-Language Models [28.416254061159176]
  Multi-modal keyphrase prediction (MMKP) aims to advance beyond text-only methods.<n>Traditional multi-modal approaches have been proven to have significant limitations in handling the challenging absence and unseen scenarios.<n>We propose leveraging vision-language models (VLMs) for the MMKP task.
  arXiv  Detail & Related papers  (2025-10-10T13:13:07Z)
• When Big Models Train Small Ones: Label-Free Model Parity Alignment for Efficient Visual Question Answering using Small VLMs [4.296395082987112]
  Large Vision-Language Models (L-VLMs) have demonstrated remarkable performance in various vision and language tasks.<n>Small Vision-Language Models (S-VLMs) offer efficiency but suffer from a significant performance gap compared to their larger counterparts.<n>We introduce the Model Parity Aligner (MPA), a novel framework designed to systematically improve S-VLMs.
  arXiv  Detail & Related papers  (2025-09-20T11:12:23Z)
• Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting [70.83781268763215]
  Vision-language models (VLMs) have achieved impressive performance across diverse multimodal tasks by leveraging large-scale pre-training.<n>VLMs face unique challenges such as cross-modal feature drift, parameter interference due to shared architectures, and zero-shot capability erosion.<n>This survey aims to serve as a comprehensive and diagnostic reference for researchers developing lifelong vision-language systems.
  arXiv  Detail & Related papers  (2025-08-06T09:03:10Z)
• Multimodal Behavioral Patterns Analysis with Eye-Tracking and LLM-Based Reasoning [12.054910727620154]
  Eye-tracking data reveals valuable insights into users' cognitive states but is difficult to analyze due to its structured, non-linguistic nature.<n>This paper presents a multimodal human-AI collaborative framework designed to enhance cognitive pattern extraction from eye-tracking signals.
  arXiv  Detail & Related papers  (2025-07-24T09:49:53Z)
• Mimicking or Reasoning: Rethinking Multi-Modal In-Context Learning in Vision-Language Models [19.361686225381447]
  Vision-language models (VLMs) are widely assumed to exhibit in-context learning (ICL)<n>We propose a new MM-ICL with Reasoning pipeline that augments each demonstration with a generated rationale alongside the answer.
  arXiv  Detail & Related papers  (2025-06-09T16:55:32Z)
• OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles [91.88062410741833]
  We introduce OpenVLThinker, one of the first open-source large vision-language models (LVLMs) to exhibit sophisticated chain-of-thought reasoning.<n>We show that OpenVLThinker-7B consistently advances performance across six benchmarks demanding mathematical and general reasoning.
  arXiv  Detail & Related papers  (2025-03-21T17:52:43Z)
• VisualPRM: An Effective Process Reward Model for Multimodal Reasoning [76.35753243272521]
  We introduce VisualPRM, which improves the reasoning abilities of existing Multimodal Large Language Models (MLLMs)<n>Our model achieves a 5.9-point improvement across seven multimodal reasoning benchmarks.<n>For the evaluation of multimodal PRMs, we propose VisualProcessBench, a benchmark with human-annotated step-wise correctness labels.
  arXiv  Detail & Related papers  (2025-03-13T12:03:37Z)
• Insight-V: Exploring Long-Chain Visual Reasoning with Multimodal Large Language Models [64.1799100754406]
  Large Language Models (LLMs) demonstrate enhanced capabilities and reliability by reasoning more.<n>Despite various efforts to improve LLM reasoning, high-quality long-chain reasoning data and optimized training pipelines still remain inadequately explored in vision-language tasks.<n>We present Insight-V, an early effort to 1) scalably produce long and robust reasoning data for complex multi-modal tasks, and 2) an effective training pipeline to enhance the reasoning capabilities of MLLMs.
  arXiv  Detail & Related papers  (2024-11-21T18:59:55Z)
• Vision-Language Instruction Tuning: A Review and Analysis [52.218690619616474]
  Vision-Language Instruction Tuning (VLIT) presents more complex characteristics compared to pure text instruction tuning. We offer a detailed categorization for existing VLIT datasets and identify the characteristics that high-quality VLIT data should possess. By incorporating these characteristics as guiding principles into the existing VLIT data construction process, we conduct extensive experiments and verify their positive impact on the performance of tuned multi-modal LLMs.
  arXiv  Detail & Related papers  (2023-11-14T14:02:32Z)
• Adapting Pre-trained Language Models to Vision-Language Tasks via Dynamic Visual Prompting [83.21164539349273]
  Pre-trained language models (PLMs) have played an increasing role in multimedia research. In this paper, we focus on exploring PLMs as a stand-alone model for vision-language reasoning tasks. We propose a novel transfer learning approach for PLMs, termed Dynamic Visual Prompting (DVP)
  arXiv  Detail & Related papers  (2023-06-01T07:19:28Z)

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.