# STAGE: Stable and Generalizable GRPO for Autoregressive Image Generation

arXiv: https://arxiv.org/abs/2509.25027 · HTML: https://mlsanity.com/paper/2509.25027/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for reasoning, one of 17 directions in LLM reasoning.
- Builds on 38 of its 54 references that are in the map.
- First posted on arXiv 29 Sep 2025.

## Placed in

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

## Builds on (most cited first, 30 of 38)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [Scaling Autoregressive Models for Content-Rich Text-to-Image Generation](https://mlsanity.com/paper/2206.10789/) (2022)
- [Chameleon: Mixed-Modal Early-Fusion Foundation Models](https://mlsanity.com/paper/2405.09818/) (2024)
- [Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis](https://mlsanity.com/paper/2306.09341/) (2023)
- [PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis](https://mlsanity.com/paper/2310.00426/) (2023)
- [Group Sequence Policy Optimization](https://mlsanity.com/paper/2507.18071/) (2025)
- [Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling](https://mlsanity.com/paper/2501.17811/) (2025)
- [Emu3: Next-Token Prediction is All You Need](https://mlsanity.com/paper/2409.18869/) (2024)
- [Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation](https://mlsanity.com/paper/2406.06525/) (2024)
- [Show-o: One Single Transformer to Unify Multimodal Understanding and Generation](https://mlsanity.com/paper/2408.12528/) (2024)
- [Flow-GRPO: Training Flow Matching Models via Online RL](https://mlsanity.com/paper/2505.05470/) (2025)
- [Vector-quantized Image Modeling with Improved VQGAN](https://mlsanity.com/paper/2110.04627/) (2021)
- [Language Model Beats Diffusion -- Tokenizer is Key to Visual Generation](https://mlsanity.com/paper/2310.05737/) (2023)
- [Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning](https://mlsanity.com/paper/2506.01939/) (2025)
- [GIT: A Generative Image-to-text Transformer for Vision and Language](https://mlsanity.com/paper/2205.14100/) (2022)
- [The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models](https://mlsanity.com/paper/2505.22617/) (2025)
- [Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation](https://mlsanity.com/paper/2410.13848/) (2024)
- [DanceGRPO: Unleashing GRPO on Visual Generation](https://mlsanity.com/paper/2505.07818/) (2025)
- [Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis](https://mlsanity.com/paper/2412.04431/) (2024)
- [T2I-R1: Reinforcing Image Generation with Collaborative Semantic-level and Token-level CoT](https://mlsanity.com/paper/2505.00703/) (2025)
- [SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer](https://mlsanity.com/paper/2501.18427/) (2025)
- [UniTok: A Unified Tokenizer for Visual Generation and Understanding](https://mlsanity.com/paper/2502.20321/) (2025)
- [Lumina-mGPT: Illuminate Flexible Photorealistic Text-to-Image Generation with Multimodal Generative Pretraining](https://mlsanity.com/paper/2408.02657/) (2024)
- [Planting a SEED of Vision in Large Language Model](https://mlsanity.com/paper/2307.08041/) (2023)
- [Can We Generate Images with CoT? Let's Verify and Reinforce Image Generation Step by Step](https://mlsanity.com/paper/2501.13926/) (2025)
- [ANOLE: An Open, Autoregressive, Native Large Multimodal Models for Interleaved Image-Text Generation](https://mlsanity.com/paper/2407.06135/) (2024)
- [Geometric-Mean Policy Optimization](https://mlsanity.com/paper/2507.20673/) (2025)

## Built on by (most cited first, 15 of 15)

- [Flow-OPD: On-Policy Distillation for Flow Matching Models](https://mlsanity.com/paper/2605.08063/) (2026)
- [Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges](https://mlsanity.com/paper/2604.13602/) (2026)
- [NextFlow: Unified Sequential Modeling Activates Multimodal Understanding and Generation](https://mlsanity.com/paper/2601.02204/) (2026)
- [Semantic Context Matters: Improving Conditioning for Autoregressive Models](https://mlsanity.com/paper/2511.14063/) (2025)
- [RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPO](https://mlsanity.com/paper/2605.15190/) (2026)
- [Learning What to Trust: Bayesian Prior-Guided Optimization for Visual Generation](https://mlsanity.com/paper/2511.18919/) (2025)
- [Embedding-perturbed Exploration Preference Optimization for Flow Models](https://mlsanity.com/paper/2605.15803/) (2026)
- [Power Reinforcement Post-Training of Text-to-Image Models with Super-Linear Advantage Shaping](https://mlsanity.com/paper/2605.10937/) (2026)
- [VAR RL Done Right: Tackling Asynchronous Policy Conflicts in Visual Autoregressive Generation](https://mlsanity.com/paper/2601.02256/) (2026)
- [Advances in GRPO for Generation Models: A Survey](https://mlsanity.com/paper/2603.06623/) (2026)
- [MAR-GRPO: Stabilized GRPO for AR-diffusion Hybrid Image Generation](https://mlsanity.com/paper/2604.06966/) (2026)
- [Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent](https://mlsanity.com/paper/2608.03979/) (2026)
- [ParaUni: Enhance Generation in Unified Multimodal Model with Reinforcement-driven Hierarchical Parallel Information Interaction](https://mlsanity.com/paper/2512.05422/) (2025)
- [B-GRTO: Bootstrapped Group Relative Tool Optimization for Referring Segmentation](https://mlsanity.com/paper/2605.23500/) (2026)
- [Policy-based Tuning of Autoregressive Image Models with Instance- and Distribution-Level Rewards](https://mlsanity.com/paper/2603.23086/) (2026)

## Abstract (arXiv)

Reinforcement learning has recently been explored to improve text-to-image generation, yet applying existing GRPO algorithms to autoregressive (AR) image models remains challenging. The instability of the training process easily disrupts the pretrained model capability during long runs, resulting in marginal gains, degraded image quality, and poor generalization. In this work, we revisit GRPO for AR image generation and identify two key issues: contradictory gradients from unnecessary tokens and unstable policy entropy dynamics. To address these, we introduce STAGE, a stable and generalizable framework that leverages two targeted solutions: 1) Advantage/KL reweighting. Similarity-aware reweighting to alleviate conflicting updates; and 2) Entropy reward. An entropy-based reward corresponding to reference model to stabilize learning. With the help of alleviating conflicts between tokens and an entropy reward for stabilizing training, we reduce disruption of the pretrained distribution and mitigate reward hacking, which in turn improves generalization and transfer better to other benchmarks. Experiments across multiple benchmarks show that STAGE consistently improves visual quality, stability, and cross-task generalization compared to baseline GRPO.