Point-RFT: Improving Multimodal Reasoning with Visually Grounded Reinforcement Finetuning

• URL: http://arxiv.org/abs/2505.19702v1
• Date: Mon, 26 May 2025 08:54:14 GMT
• Title: Point-RFT: Improving Multimodal Reasoning with Visually Grounded Reinforcement Finetuning
• Authors: Minheng Ni, Zhengyuan Yang, Linjie Li, Chung-Ching Lin, Kevin Lin, Wangmeng Zuo, Lijuan Wang,
• Abstract summary: We introduce Point-RFT, a multimodal reasoning framework explicitly designed to leverage visually grounded CoT reasoning for visual document understanding.<n>Our approach consists of two stages: First, we conduct format finetuning using a curated dataset of 71K diverse visual reasoning problems, each annotated with detailed, step-by-step rationales explicitly grounded to corresponding visual elements.<n>On ChartQA, our approach improves accuracy from 70.88% (language-finetuned baseline) to 90.04%, surpassing the 83.92% accuracy achieved by reinforcement finetuning relying solely on text-based CoT.
• Score: 122.81815833343026
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Recent advances in large language models have significantly improved textual reasoning through the effective use of Chain-of-Thought (CoT) and reinforcement learning. However, extending these successes to vision-language tasks remains challenging due to inherent limitations in text-only CoT, such as visual hallucinations and insufficient multimodal integration. In this paper, we introduce Point-RFT, a multimodal reasoning framework explicitly designed to leverage visually grounded CoT reasoning for visual document understanding. Our approach consists of two stages: First, we conduct format finetuning using a curated dataset of 71K diverse visual reasoning problems, each annotated with detailed, step-by-step rationales explicitly grounded to corresponding visual elements. Second, we employ reinforcement finetuning targeting visual document understanding. On ChartQA, our approach improves accuracy from 70.88% (format-finetuned baseline) to 90.04%, surpassing the 83.92% accuracy achieved by reinforcement finetuning relying solely on text-based CoT. The result shows that our grounded CoT is more effective for multimodal reasoning compared with the text-only CoT. Moreover, Point-RFT exhibits superior generalization capability across several out-of-domain visual document reasoning benchmarks, including CharXiv, PlotQA, IconQA, TabMWP, etc., and highlights its potential in complex real-world scenarios.

Related papers

• Multimodal Fact Checking with Unified Visual, Textual, and Contextual Representations [2.139909491081949]
  We propose a unified framework for fine-grained multimodal fact verification called "MultiCheck"<n>Our architecture combines dedicated encoders for text and images with a fusion module that captures cross-modal relationships using element-wise interactions.<n>We evaluate our approach on the Factify 2 dataset, achieving a weighted F1 score of 0.84, substantially outperforming the baseline.
  arXiv  Detail & Related papers  (2025-08-07T07:36:53Z)
• Decouple before Align: Visual Disentanglement Enhances Prompt Tuning [85.91474962071452]
  Prompt tuning (PT) has showcased remarkable effectiveness in improving the task-specific transferability of vision-language models.<n>This paper delves into a previously overlooked information asymmetry issue in PT, where the visual modality mostly conveys more context.<n>We propose DAPT, an effective PT framework based on an intuitive decouple-before-align concept.
  arXiv  Detail & Related papers  (2025-08-01T07:46:00Z)
• Can Generated Images Serve as a Viable Modality for Text-Centric Multimodal Learning? [3.966028515034415]
  This work systematically investigates whether images generated on-the-fly by Text-to-Image (T2I) models can serve as a valuable complementary modality for text-centric tasks.
  arXiv  Detail & Related papers  (2025-06-21T07:32:09Z)
• Reinforcing Video Reasoning with Focused Thinking [65.85683941058916]
  We propose TW-GRPO, a novel framework that enhances visual reasoning with focused thinking and dense reward granularity.<n>Specifically, we employ a token weighting mechanism that prioritizes tokens with high informational density.<n>We also reformulate RL training by shifting from single-choice to multi-choice QA tasks.
  arXiv  Detail & Related papers  (2025-05-30T15:42:19Z)
• Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning [19.28434717501445]
  Visual reasoning abilities play a crucial role in understanding complex multimodal data.<n>Existing methods improve VLM reasoning via Chain-of-Thought supervised fine-tuning.<n>We propose Reason-RFT, a novel reinforcement fine-tuning framework.
  arXiv  Detail & Related papers  (2025-03-26T17:38:06Z)
• 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)
• Separate-and-Enhance: Compositional Finetuning for Text2Image Diffusion Models [58.46926334842161]
  This work illuminates the fundamental reasons for such misalignment, pinpointing issues related to low attention activation scores and mask overlaps. We propose two novel objectives, the Separate loss and the Enhance loss, that reduce object mask overlaps and maximize attention scores. Our method diverges from conventional test-time-adaptation techniques, focusing on finetuning critical parameters, which enhances scalability and generalizability.
  arXiv  Detail & Related papers  (2023-12-10T22:07:42Z)
• Vision-Enhanced Semantic Entity Recognition in Document Images via Visually-Asymmetric Consistency Learning [19.28860833813788]
  Existing models commonly train a visual encoder with weak cross-modal supervision signals. We propose a novel textbfVisually-textbfAsymmetric cotextbfNsistentextbfCy textbfLearning (textscVancl) approach to capture fine-grained visual and layout features.
  arXiv  Detail & Related papers  (2023-10-23T10:37:22Z)
• PV2TEA: Patching Visual Modality to Textual-Established Information Extraction [59.76117533540496]
  We patch the visual modality to the textual-established attribute information extractor. PV2TEA is an encoder-decoder architecture equipped with three bias reduction schemes. Empirical results on real-world e-Commerce datasets demonstrate up to 11.74% absolute (20.97% relatively) F1 increase over unimodal baselines.
  arXiv  Detail & Related papers  (2023-06-01T05:39:45Z)
• Re-evaluating the Need for Multimodal Signals in Unsupervised Grammar Induction [119.07444667700473]
  Recent work has shown that multimodal training inputs can improve grammar induction. We design a stronger text-only baseline, which we refer to as LC-PCFG. LC-PCFG provides an up to 17% relative improvement in Corpus-F1 compared to state-of-the-art multimodal grammar induction methods.
  arXiv  Detail & Related papers  (2022-12-20T18:59:50Z)
• Cosine meets Softmax: A tough-to-beat baseline for visual grounding [17.316608734530124]
  Our framework minimizes the cross-entropy loss over the cosine distance between multiple image ROI features with a text embedding. We perform experiments on the Talk2Car dataset and achieve 68.7% AP50 accuracy, improving upon the previous state of the art by 8.6%.
  arXiv  Detail & Related papers  (2020-09-13T19:35:43Z)

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.