# SpatialFlow-GRPO: Where Spatial Credit Drives Image Editing

arXiv: https://arxiv.org/abs/2606.26872 · HTML: https://mlsanity.com/paper/2606.26872/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning.
- Builds on 32 of its 51 references that are in the map.
- First posted on arXiv 25 Jun 2026.

## Placed in

- LLM reasoning / [Reinforcement learning for reasoning](https://mlsanity.com/lr/reinforcement-learning-for-reasoning/)

## Builds on (most cited first, 30 of 32)

- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Prompt-to-Prompt Image Editing with Cross Attention Control](https://mlsanity.com/paper/2208.01626/) (2022)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations](https://mlsanity.com/paper/2108.01073/) (2021)
- [Training Diffusion Models with Reinforcement Learning](https://mlsanity.com/paper/2305.13301/) (2023)
- [Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis](https://mlsanity.com/paper/2306.09341/) (2023)
- [Flow-GRPO: Training Flow Matching Models via Online RL](https://mlsanity.com/paper/2505.05470/) (2025)
- [DiffEdit: Diffusion-based semantic image editing with mask guidance](https://mlsanity.com/paper/2210.11427/) (2022)
- [Step1X-Edit: A Practical Framework for General Image Editing](https://mlsanity.com/paper/2504.17761/) (2025)
- [DanceGRPO: Unleashing GRPO on Visual Generation](https://mlsanity.com/paper/2505.07818/) (2025)
- [OmniGen2: Towards Instruction-Aligned Multimodal Generation](https://mlsanity.com/paper/2506.18871/) (2025)
- [ImgEdit: A Unified Image Editing Dataset and Benchmark](https://mlsanity.com/paper/2505.20275/) (2025)
- [VIEScore: Towards Explainable Metrics for Conditional Image Synthesis Evaluation](https://mlsanity.com/paper/2312.14867/) (2023)
- [HQ-Edit: A High-Quality Dataset for Instruction-based Image Editing](https://mlsanity.com/paper/2404.09990/) (2024)
- [Guiding Instruction-based Image Editing via Multimodal Large Language Models](https://mlsanity.com/paper/2309.17102/) (2023)
- [EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling](https://mlsanity.com/paper/2509.23909/) (2025)
- [EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing](https://mlsanity.com/paper/2509.26346/) (2025)
- [SeedEdit: Align Image Re-Generation to Image Editing](https://mlsanity.com/paper/2411.06686/) (2024)
- [Multimodal RewardBench 2: Evaluating Omni Reward Models for Interleaved Text and Image](https://mlsanity.com/paper/2512.16899/) (2025)
- [InstructRL4Pix: Training Diffusion for Image Editing by Reinforcement Learning](https://mlsanity.com/paper/2406.09973/) (2024)
- [ImageDoctor: Diagnosing Text-to-Image Generation via Grounded Image Reasoning](https://mlsanity.com/paper/2510.01010/) (2025)
- [SpatialReward: Bridging the Perception Gap in Online RL for Image Editing via Explicit Spatial Reasoning](https://mlsanity.com/paper/2602.07458/) (2026)
- [HP-Edit: A Human-Preference Post-Training Framework for Image Editing](https://mlsanity.com/paper/2604.19406/) (2026)

## Built on by (most cited first, 2 of 2)

- [Sprout: Building Dynamic Memory While Reasoning for Agentic Video Understanding](https://mlsanity.com/paper/2609.35497/) (2026)
- [LoGo: Local-Global Rewards for Consistent Long-Horizon Video Generation](https://mlsanity.com/paper/2610.03636/) (2026)

## Abstract (arXiv)

Recent online reinforcement learning has substantially improved image editing quality. However, existing Flow-GRPO-style methods usually rely on a single whole-image reward, which makes fine-grained editing optimization difficult. We observe that a key obstacle in image editing is this spatial uniformity assumption: a whole-image reward cannot distinguish how different spatial regions contribute to image quality. To address this issue, we propose SpatialFlow-GRPO, a training framework that introduces spatially fine-grained reward feedback. The framework converts region-aware rewards into semantic-region-level optimization signals and aligns region advantages with the corresponding latent positions during policy updates. We also train a region-aware reward model, SFReward, construct SFReward-14K with region-annotated editing samples, and introduce MultiEditBench to evaluate multi-region editing ability. On OmniGen2 and FLUX.2-klein-4B, SpatialFlow-GRPO outperforms Flow-GRPO on GEdit-Bench, ImgEdit-Bench, and MultiEditBench. The results show that SpatialFlow-GRPO converts local feedback into spatially aligned update signals and improves editing quality.