mlsanityData to 14 Oct 2026

Ground-R1: Incentivizing Grounded Visual Reasoning via Reinforcement Learning

arXiv 2505.20272 ↗First posted 26 May 2025cs.CVcited by 49

On the map: LLM reasoning → Multimodal reasoning

also in 1 more
Abstract (arXiv)

Large Vision-Language Models (LVLMs) have become powerful general-purpose assistants, yet their predictions often lack reliability and interpretability due to insufficient grounding in visual evidence. The emerging thinking-with-images paradigm seeks to address this issue by explicitly anchoring reasoning to image regions. However, we empirically find that most existing methods suffer from a systematic scale-driven bias in optimization, where training rewards are dominated by large visual regions, suppressing learning from small but semantically critical evidence and leading to spurious grounding at inference time. To address this limitation, we propose Ground-R1, a de-biased thinking-with-images framework trained via a novel Scale Relative Policy Optimization (SRPO) objective that replaces standard GRPO. Specifically, our SRPO recalibrates reward learning across evidence regions of different sizes through scale-aware binning and intra-/inter-bin comparisons, enabling balanced credit assignment during training. Experimental results on general LVLM, high-resolution, and visual grounding benchmarks validate the effectiveness of Ground-R1 and show that SRPO yields consistent gains over standard GRPO in both response accuracy and evidence grounding.

Key facts. Cited by 49 papers in the map (data through 14 Oct 2026). Placed in Multimodal reasoning, one of 17 directions in LLM reasoning. Builds on 72 of its 89 references that are in the map. First posted on arXiv 26 May 2025.

Builds on

72 of 89 references.

Closest ancestors · 26

Same direction, newest first

  1. 2025Pixel Reasoner: Incentivizing Pixel-Space Reasoning with Curiosity-Driven Reinforcement Learning21 May
  2. 2025DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning20 May
  3. 2025OpenThinkIMG: Learning to Think with Images via Visual Tool Reinforcement Learning13 May
  4. 2025Perception-R1: Pioneering Perception Policy with Reinforcement Learning10 Apr
  5. 2025SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-Improvement10 Apr
  6. 2025Grounded Chain-of-Thought for Multimodal Large Language Models17 Mar
  7. 2025Multimodal Chain-of-Thought Reasoning: A Comprehensive Survey16 Mar
  8. 2025R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization13 Mar
  9. 2025LMM-R1: Empowering 3B LMMs with Strong Reasoning Abilities Through Two-Stage Rule-Based RL10 Mar
  10. 2025MM-Eureka: Exploring the Frontiers of Multimodal Reasoning with Rule-based Reinforcement Learning10 Mar
  11. 2025Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models9 Mar
  12. 2025Visual-RFT: Visual Reinforcement Fine-Tuning3 Mar
14 more
  1. 2025LlamaV-o1: Rethinking Step-by-step Visual Reasoning in LLMs10 Jan
  2. 2025ReFocus: Visual Editing as a Chain of Thought for Structured Image Understanding9 Jan
  3. 2024Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling6 Dec
  4. 2024Insight-V: Exploring Long-Chain Visual Reasoning with Multimodal Large Language Models21 Nov
  5. 2024LLaVA-CoT: Let Vision Language Models Reason Step-by-Step15 Nov
  6. 2024Visual Agents as Fast and Slow Thinkers16 Aug
  7. 2024Visual Sketchpad: Sketching as a Visual Chain of Thought for Multimodal Language Models13 Jun
  8. 2024Visual CoT: Advancing Multi-Modal Language Models with a Comprehensive Dataset and Benchmark for Chain-of-Thought Reasoning25 Mar
  9. 2024Chain-of-Spot: Interactive Reasoning Improves Large Vision-Language Models19 Mar
  10. 2023T-SciQ: Teaching Multimodal Chain-of-Thought Reasoning via Mixed Large Language Model Signals for Science Question Answering5 May
  11. 2023ViperGPT: Visual Inference via Python Execution for Reasoning14 Mar
  12. 2023Multimodal Chain-of-Thought Reasoning in Language Models2 Feb
  13. 2023See, Think, Confirm: Interactive Prompting Between Vision and Language Models for Knowledge-based Visual Reasoning12 Jan
  14. 2022Visual Programming: Compositional visual reasoning without training18 Nov

Other references · 46

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  3. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  4. 2025Qwen2.5-VL Technical Reportcited by 5,587
  5. 2023Improved Baselines with Visual Instruction Tuningcited by 4,418
  6. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  7. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,332
  8. 2024Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextcited by 3,224
  9. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,211
  10. 2024LLaVA-OneVision: Easy Visual Task Transfercited by 2,719
  11. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,729
  12. 2019Towards VQA Models That Can Readcited by 1,718
34 more
  1. 2023MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Modelscited by 1,652
  2. 2015Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Modelscited by 1,362
  3. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,317
  4. 2023MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilitiescited by 1,239
  5. 2023Kosmos-2: Grounding Multimodal Large Language Models to the Worldcited by 1,089
  6. 2024How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suitescited by 1,050
  7. 2023SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehensioncited by 940
  8. 2024Chameleon: Mixed-Modal Early-Fusion Foundation Modelscited by 927
  9. 2023Shikra: Unleashing Multimodal LLM's Referential Dialogue Magiccited by 906
  10. 2015Generation and Comprehension of Unambiguous Object Descriptionscited by 904
  11. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 886
  12. 2023CogVLM: Visual Expert for Pretrained Language Modelscited by 708
  13. 2022Towards Reasoning in Large Language Models: A Surveycited by 596
  14. 2023V*: Guided Visual Search as a Core Mechanism in Multimodal LLMscited by 482
  15. 2023Generative Multimodal Models are In-Context Learnerscited by 446
  16. 2021InfographicVQAcited by 437
  17. 2020TextCaps: a Dataset for Image Captioning with Reading Comprehensioncited by 400
  18. 2015Visual7W: Grounded Question Answering in Imagescited by 318
  19. 2022Visual Spatial Reasoningcited by 293
  20. 2025Towards Large Reasoning Models: A Survey of Reinforced Reasoning with Large Language Modelscited by 293
  21. 2024MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?cited by 292
  22. 2023SPHINX: The Joint Mixing of Weights, Tasks, and Visual Embeddings for Multi-modal Large Language Modelscited by 286
  23. 2021ICDAR2019 Competition on Scanned Receipt OCR and Information Extractioncited by 256
  24. 2024ReFT: Reasoning with Reinforced Fine-Tuningcited by 239
  25. 2023Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMscited by 199
  26. 2023LLaVA-Plus: Learning to Use Tools for Creating Multimodal Agentscited by 193
  27. 2025MLLMs Know Where to Look: Training-free Perception of Small Visual Details with Multimodal LLMscited by 176
  28. 2024Mulberry: Empowering MLLM with o1-like Reasoning and Reflection via Collective Monte Carlo Tree Searchcited by 175
  29. 2023Document Understanding Dataset and Evaluation (DUDE)cited by 135
  30. 2024Pride and Prejudice: LLM Amplifies Self-Bias in Self-Refinementcited by 114
  31. 2025Visual Planning: Let's Think Only with Imagescited by 85
  32. 2024VoCoT: Unleashing Visually Grounded Multi-Step Reasoning in Large Multi-Modal Modelscited by 43
  33. 2024CAD-Assistant: Tool-Augmented VLLMs as Generic CAD Task Solverscited by 40
  34. 2024UniBias: Unveiling and Mitigating LLM Bias through Internal Attention and FFN Manipulationcited by 31

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 17; 2026: 32.

49 papers.

Sort:

By area:

  1. 2025The Landscape of Agentic Reinforcement Learning for LLMs: A Surveycited by 205
  2. 2025A Survey of Reinforcement Learning for Large Reasoning Modelscited by 180
  3. 2025Traceable Evidence Enhanced Visual Grounded Reasoning: Evaluation and Methodologycited by 69
  4. 2025Explain Before You Answer: A Survey on Compositional Visual Reasoningcited by 28
  5. 2026Embed-RL: Reinforcement Learning for Reasoning-Driven Multimodal Embeddingscited by 24
  6. 2026LensWalk: Agentic Video Understanding by Planning How You See in Videoscited by 17
  7. 2026Imagination Helps Visual Reasoning, But Not Yet in Latent Spacecited by 12
  8. 2025DeFacto: Counterfactual Thinking with Images for Enforcing Evidence-Grounded and Faithful Reasoningcited by 10
  9. 2025textscGUI-Spotlight: Adaptive Iterative Focus Refinement for Enhanced GUI Visual Groundingcited by 10
  10. 2025Reinforcement Learning for Large Model: A Surveycited by 9
  11. 2026Seeing with You: Perception-Reasoning Coevolution for Multimodal Reasoningcited by 9
  12. 2026CreatiParser: Generative Image Parsing of Raster Graphic Designs into Editable Layerscited by 9
37 more
  1. 2025VG-Refiner: Towards Tool-Refined Referring Grounded Reasoning via Agentic Reinforcement Learningcited by 8
  2. 2026Staying VIGILant: Mitigating Visual Laziness via Counterfactual Visual Alignment in MLLMscited by 8
  3. 2025PhyBlock: A Progressive Benchmark for Physical Understanding and Planning via 3D Block Assemblycited by 7
  4. 2025Evidence-Grounded Trustworthy Multimodal Reasoning and Evaluation Benchmark in Complex Urban Scenescited by 7
  5. 2025Mixture-of-Visual-Thoughts: Exploring Context-Adaptive Reasoning Mode Selection for General Visual Reasoningcited by 7
  6. 2025Video Spatial Reasoning with Object-Centric 3D Rolloutcited by 6
  7. 2026Can Vision Language Models Assess Graphic Design Aesthetics? A Benchmark, Evaluation, and Dataset Perspectivecited by 6
  8. 2026DRS-GUI: Dynamic Region Search for Training-Free GUI Groundingcited by 5
  9. 2025Visual serial processing deficits explain divergences in human and VLM reasoningcited by 4
  10. 2025Training Multi-Image Vision Agents via End2End Reinforcement Learningcited by 4
  11. 2026OddGridBench: Exposing the Lack of Fine-Grained Visual Discrepancy Sensitivity in Multimodal Large Language Modelscited by 4
  12. 2026Annotations as Rollouts: Efficient and Scalable Reinforcement Learning for Video MLLMscited by 4
  13. 2025Look as You Think: Unifying Reasoning and Visual Evidence Attribution for Verifiable Document RAG via Reinforcement Learningcited by 3
  14. 2026Disentangling Perception and Reasoning in Multimodal LLMs via Reward Designcited by 3
  15. 2026APPO: Attention-guided Perception Policy Optimization for Video Reasoningcited by 3
  16. 2026When Thinking Hurts: Mitigating Visual Forgetting in Video Reasoning via Frame Repetitioncited by 3
  17. 2026From Web to Pixels: Bringing Agentic Search into Visual Perceptioncited by 3
  18. 2025DiVE-k: Differential Visual Reasoning for Fine-grained Image Recognitioncited by 2
  19. 2025Thinking with Drafts: Speculative Temporal Reasoning for Efficient Long Video Understandingcited by 2
  20. 2026Structured Role-Aware Policy Optimization for Multimodal Reasoningcited by 2
  21. 2026When RAG Hurts: Diagnosing and Mitigating Attention Distraction in Retrieval-Augmented LVLMscited by 1
  22. 2026Faithful-MR1: Faithful Multimodal Reasoning via Anchoring and Reinforcing Visual Attentioncited by 1
  23. 2026Iterative Visual Thinking and the Self-Correction Mirage in VLM Groundingcited by 1
  24. 2026Actor as Its Own Critic: Unifying Region Understanding and Localization via CycleGRPOcited by 1
  25. 2026Graph-PiT: Enhancing Structural Coherence in Part-Based Image Synthesis via Graph Priorscited by 0
  26. 2026GazeVLM: Active Vision via Internal Attention Control for Multimodal Reasoningcited by 0
  27. 2026iVGR: Internalizing Visually Grounded Reasoning for MLLMs with Reinforcement Learningcited by 0
  28. 2026An LMM for Precisely Grounding Elements in Documentscited by 0
  29. 2026SER: Learning to Ground Video Reasoning with Semantic Evidence Rewardscited by 0
  30. 2026From Structure to Synergy: A Survey of Vision-Language Perception Paradigm Evolution in Multimodal Large Language Modelscited by 0
  31. 2026Reasoning-Guided Part-Level Visual Grounding via Reinforcement Learningcited by 0
  32. 2026Evidence-Grounded Trustworthy Multimodal Reasoning and Evaluation Benchmark in Complex Urban Scenescited by 0
  33. 2026DocMIDE: Learning Multi-Hop Implicit Derivation in Visually Rich Documentscited by 0
  34. 2026From Reward Signal to Visual Utility: A Controlled Audit of Medical VLM Post-Trainingcited by 0
  35. 2026Answer with Evidence: Consistency-Aware Grounded Visual Question Answering for Roadside Traffic Scenescited by 0
  36. 2026Thinking with Images without Coordinatescited by 0
  37. 2026Adaptive Look, Flexible Thought: Learning How and How Long to Think in Multimodal Reasoningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025UniVG-R1: Reasoning Guided Universal Visual Grounding with Reinforcement Learning
  2. 2025Grounded Reinforcement Learning for Visual Reasoning
  3. 2026Thinking with Visual Grounding
  4. 2026H-GRPO: Permutation-Invariant Reinforcement Learning for Grounded Visual Reasoning
  5. 2025MIRG-RL: Multi-Image Reasoning and Grounding with Reinforcement Learning
  6. 2026MG-Thinker: Bi-Axial Self-Reflection for Multi-Image Reasoning Grounding
  7. 2025High-Resolution Visual Reasoning via Multi-Turn Grounding-Based Reinforcement Learning
  8. 2026Not All Tokens See Equally: Perception-Grounded Policy Optimization for Large Vision-Language Models