Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models

• URL: http://arxiv.org/abs/2503.20752v3
• Date: Sun, 05 Oct 2025 00:51:14 GMT
• Title: Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models
• Authors: Huajie Tan, Yuheng Ji, Xiaoshuai Hao, Xiansheng Chen, Pengwei Wang, Zhongyuan Wang, Shanghang Zhang,
• Abstract summary: Reason-RFT is a two-stage reinforcement fine-tuning framework for visual reasoning.<n>First,Supervised Fine-Tuning (SFT) with curated CoT data activates the reasoning potential of Vision-Language Models (VLMs)<n>Second, reinforcement learning based on Group Relative Policy Optimization (GRPO) generates multiple reasoning-response pairs to enhance adaptability to domain shifts.
• Score: 42.75418134743927
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Visual reasoning abilities play a crucial role in understanding complex multimodal data, advancing both domain-specific applications and artificial general intelligence (AGI). Existing methods enhance Vision-Language Models (VLMs) through Chain-of-Thought (CoT) supervised fine-tuning using meticulously annotated data. However, this approach may lead to overfitting and cognitive rigidity, limiting the model's generalization ability under domain shifts and reducing real-world applicability. To overcome these limitations, we propose Reason-RFT, a two-stage reinforcement fine-tuning framework for visual reasoning. First, Supervised Fine-Tuning (SFT) with curated CoT data activates the reasoning potential of VLMs. This is followed by reinforcement learning based on Group Relative Policy Optimization (GRPO), which generates multiple reasoning-response pairs to enhance adaptability to domain shifts. To evaluate Reason-RFT, we reconstructed a comprehensive dataset covering visual counting, structural perception, and spatial transformation, serving as a benchmark for systematic assessment across three key dimensions. Experimental results highlight three advantages: (1) performance enhancement, with Reason-RFT achieving state-of-the-art results and outperforming both open-source and proprietary models; (2) generalization superiority, maintaining robust performance under domain shifts across various tasks; and (3) data efficiency, excelling in few-shot learning scenarios and surpassing full-dataset SFT baselines. Reason-RFT introduces a novel training paradigm for visual reasoning and marks a significant step forward in multimodal research. Project website: https://tanhuajie.github.io/ReasonRFT

Related papers

• Long Grounded Thoughts: Distilling Compositional Visual Reasoning Chains at Scale [70.23466957404891]
  We introduce a new reasoning data generation framework spanning diverse skills and levels of complexity with over 1M high-quality synthetic vision-centric questions.<n>We show that finetuning Qwen2.5-VL-7B on our data outperforms all open-data baselines across all evaluated vision-centric benchmarks.
  arXiv  Detail & Related papers  (2025-11-07T20:50:54Z)
• CIR-CoT: Towards Interpretable Composed Image Retrieval via End-to-End Chain-of-Thought Reasoning [93.05917922306196]
  Composed Image Retrieval (CIR) aims to find a target image from a reference image and a modification text.<n>CIR-CoT is the first end-to-end retrieval-oriented MLLM designed to integrate explicit Chain-of-Thought (CoT) reasoning.
  arXiv  Detail & Related papers  (2025-10-09T09:41:45Z)
• Simple o3: Towards Interleaved Vision-Language Reasoning [38.46230601239066]
  We propose Simple o3, an end-to-end framework that integrates dynamic tool interactions into interleaved vision-language reasoning.<n>Our approach features a scalable data synthesis pipeline that generates high-quality interleaved vision-language reasoning chains.<n> Experimental results demonstrate Simple o3's superior performance on diverse benchmarks, outperforming existing approaches.
  arXiv  Detail & Related papers  (2025-08-16T17:15:39Z)
• Direct Reasoning Optimization: LLMs Can Reward And Refine Their Own Reasoning for Open-Ended Tasks [6.881699020319577]
  We propose Direct Reasoning Optimization (DRO), a reinforcement learning framework for fine-tuning Large Language Models (LLMs)<n>DRO is guided by a new reward signal: the Reasoning Reflection Reward (R3)<n>DRO consistently outperforms strong baselines while remaining broadly applicable across both open-ended and structured domains.
  arXiv  Detail & Related papers  (2025-06-16T10:43:38Z)
• Learning Efficient and Generalizable Graph Retriever for Knowledge-Graph Question Answering [75.12322966980003]
  Large Language Models (LLMs) have shown strong inductive reasoning ability across various domains.<n>Most existing RAG pipelines rely on unstructured text, limiting interpretability and structured reasoning.<n>Recent studies have explored integrating knowledge graphs with LLMs for knowledge graph question answering.<n>We propose RAPL, a novel framework for efficient and effective graph retrieval in KGQA.
  arXiv  Detail & Related papers  (2025-06-11T12:03:52Z)
• WeThink: Toward General-purpose Vision-Language Reasoning via Reinforcement Learning [17.459985667824807]
  Building on the success of text-based reasoning models like DeepSeek-R1, extending these capabilities to multimodal reasoning holds great promise.<n>In this paper, we show how to achieve the general-purpose visual-language reasoning through reinforcement learning.
  arXiv  Detail & Related papers  (2025-06-09T16:20:54Z)
• Perceptual Decoupling for Scalable Multi-modal Reasoning via Reward-Optimized Captioning [78.17782197231325]
  We propose a reasoning-guided reinforcement learning strategy that aligns the extractor's captioning behavior with the reasoning objective.<n> Experiments on multi-modal math and science benchmarks show that the proposed RACRO method achieves state-of-the-art average performance.
  arXiv  Detail & Related papers  (2025-06-05T02:28:07Z)
• Unveiling the Compositional Ability Gap in Vision-Language Reasoning Model [39.58344147240552]
  We investigate whether large vision-language models (VLMs) can compose capabilities across modalities or tasks under out-of-distribution conditions.<n>Our findings shed light on the current limitations of RL-based reasoning VLM training and provide actionable insights toward building models that reason compositionally across modalities and tasks.
  arXiv  Detail & Related papers  (2025-05-26T01:42:38Z)
• VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model [29.524164786422368]
  Recently, DeepSeek R1 has shown that reinforcement learning can substantially improve the reasoning capabilities of Large Language Models (LLMs) We investigate the extension of R1-style reinforcement learning to Vision-Language Models (VLMs) We develop VLM-R1, a dedicated framework designed to harness RL for improving VLMs' performance on general vision-language tasks.
  arXiv  Detail & Related papers  (2025-04-10T10:05:15Z)
• Exploring the Effect of Reinforcement Learning on Video Understanding: Insights from SEED-Bench-R1 [53.894789613838654]
  We introduce SEED-Bench-R1, a benchmark designed to evaluate post-training methods for MLLMs in video understanding. It includes intricate real-world videos and complex everyday planning tasks in the format of multiple-choice questions. Using Qwen2-VL-Instruct-7B as a base model, we compare RL with supervised fine-tuning (SFT) Our detailed analysis reveals that RL enhances visual perception but often produces less coherent reasoning chains.
  arXiv  Detail & Related papers  (2025-03-31T17:55:23Z)
• OpenVLThinker: An Early Exploration to Complex Vision-Language Reasoning via Iterative Self-Improvement [91.88062410741833]
  This study investigates whether similar reasoning capabilities can be successfully integrated into large vision-language models (LVLMs)<n>We consider an approach that iteratively leverages supervised fine-tuning (SFT) on lightweight training data and Reinforcement Learning (RL) to further improve model generalization.<n>OpenVLThinker, a LVLM exhibiting consistently improved reasoning performance on challenging benchmarks such as MathVista, MathVerse, and MathVision, demonstrates the potential of our strategy for robust vision-language reasoning.
  arXiv  Detail & Related papers  (2025-03-21T17:52:43Z)
• R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization [26.757458496178437]
  We introduce R1-Onevision, a multimodal reasoning model designed to bridge the gap between visual perception and deep reasoning.<n>We construct the R1-Onevision dataset which provides detailed, step-by-step multimodal reasoning annotations across diverse domains.<n>We further develop the R1-Onevision model through supervised fine-tuning and reinforcement learning to cultivate advanced reasoning.<n> Experimental results show that R1-Onevision achieves state-of-the-art performance, outperforming models such as GPT-4o and Qwen2.5-VL.
  arXiv  Detail & Related papers  (2025-03-13T17:56:05Z)
• Visual-RFT: Visual Reinforcement Fine-Tuning [75.20572976629646]
  Reinforcement Fine-Tuning (RFT) in Large Reasoning Models like OpenAI o1 learns from feedback on its answers.<n>Visual-RFT further extends the application areas of RFT on visual tasks.
  arXiv  Detail & Related papers  (2025-03-03T18:16:32Z)
• Cognitive Paradigms for Evaluating VLMs on Visual Reasoning Task [3.2228025627337864]
  Advancing machine visual reasoning requires a deeper understanding of how Vision-Language Models (VLMs) process and interpret complex visual patterns.<n>This work introduces a novel, cognitively-inspired evaluation framework to systematically analyze VLM reasoning on natural image-based Bongard Problems.
  arXiv  Detail & Related papers  (2025-01-23T12:42:42Z)
• DRIVINGVQA: Analyzing Visual Chain-of-Thought Reasoning of Vision Language Models in Real-World Scenarios with Driving Theory Tests [69.00444996464662]
  We present DrivingVQA, a new benchmark derived from driving theory tests to evaluate visual chain-of-thought reasoning in complex real-world scenarios.<n>Our experiments reveal that open-source and proprietary LVLMs struggle with visual chain-of-thought reasoning under zero-shot settings.<n>We investigate training strategies that leverage relevant entities to improve visual reasoning.
  arXiv  Detail & Related papers  (2025-01-08T18:31:16Z)
• STEP: Enhancing Video-LLMs' Compositional Reasoning by Spatio-Temporal Graph-guided Self-Training [87.58996020705258]
  Video Large Language Models (Video-LLMs) have recently shown strong derivation in basic video understanding tasks.<n>Video-LLMs struggle with compositional reasoning that requires multi-step explicit-temporal inference across object relations, interactions and events.<n>We propose STEP, a novel graph-guided self-training method that enables VideoLLMs to generate reasoning-rich finetuning data from any raw videos to improve itself.
  arXiv  Detail & Related papers  (2024-11-29T11:54:55Z)
• Improve Vision Language Model Chain-of-thought Reasoning [86.83335752119741]
  Chain-of-thought (CoT) reasoning in vision language models (VLMs) is crucial for improving interpretability and trustworthiness. We show that training VLM on short answers does not generalize well to reasoning tasks that require more detailed responses.
  arXiv  Detail & Related papers  (2024-10-21T17:00:06Z)
• Rephrase and Contrast: Fine-Tuning Language Models for Enhanced Understanding of Communication and Computer Networks [13.829525575305206]
  This paper introduces our Rephrase and Contrast (RaC) framework, an efficient fine-tuning framework. RaC enhances LLMs' comprehension and critical thinking abilities by incorporating question reformulation and contrastive analysis. To efficiently construct the dataset for RaC fine-tuning, we develop a GPT-assisted data mining method for generating high-quality question-answer pairs.
  arXiv  Detail & Related papers  (2024-09-21T16:04:43Z)
• Cantor: Inspiring Multimodal Chain-of-Thought of MLLM [83.6663322930814]
  We argue that converging visual context acquisition and logical reasoning is pivotal for tackling visual reasoning tasks. We propose an innovative multimodal CoT framework, termed Cantor, characterized by a perception-decision architecture. Our experiments demonstrate the efficacy of the proposed framework, showing significant improvements in multimodal CoT performance.
  arXiv  Detail & Related papers  (2024-04-24T17:59:48Z)

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.