English

Smart-GRPO: Smartly Sampling Noise for Efficient RL of Flow-Matching Models

Computer Vision and Pattern Recognition 2025-10-06 v1

Abstract

Recent advancements in flow-matching have enabled high-quality text-to-image generation. However, the deterministic nature of flow-matching models makes them poorly suited for reinforcement learning, a key tool for improving image quality and human alignment. Prior work has introduced stochasticity by perturbing latents with random noise, but such perturbations are inefficient and unstable. We propose Smart-GRPO, the first method to optimize noise perturbations for reinforcement learning in flow-matching models. Smart-GRPO employs an iterative search strategy that decodes candidate perturbations, evaluates them with a reward function, and refines the noise distribution toward higher-reward regions. Experiments demonstrate that Smart-GRPO improves both reward optimization and visual quality compared to baseline methods. Our results suggest a practical path toward reinforcement learning in flow-matching frameworks, bridging the gap between efficient training and human-aligned generation.

Keywords

Cite

@article{arxiv.2510.02654,
  title  = {Smart-GRPO: Smartly Sampling Noise for Efficient RL of Flow-Matching Models},
  author = {Benjamin Yu and Jackie Liu and Justin Cui},
  journal= {arXiv preprint arXiv:2510.02654},
  year   = {2025}
}