外观
论文arXiv 2610.1627516 天前
强化学习中的 Flow Matching 综述
原文链接：A Survey of Flow Matching in Reinforcement Learning（arXiv 2610.16275 · 发布日期：2026-10-12）
A Survey of Flow Matching in Reinforcement Learning
本文针对强化学习中 Flow Matching 文献碎片化问题，按建模目标和 RL 信号整合机制梳理相关方法。
核心要点
- 动机是 Flow Matching 正在强化学习中快速发展，但离线控制、在线微调和对齐等方向文献分散。
- 方法上按目标分布和 RL 信号整合机制建立分类，包括策略、价值、动力学及多类优化方式。
- 综述比较 Flow Matching 与高斯、扩散基线，并讨论离散与非欧动作空间等新兴前沿。
- 论文旨在提供生成式强化学习与对齐的路线图，并概述关键开放问题。
验证范围与局限：本文是综述，范围覆盖离线连续控制、在线微调和基础模型对齐等方向，摘要未报告具体实机或仿真实验。
英文摘要 (Abstract)
Flow Matching (FM) has recently emerged as a principled and efficient generative modeling framework for reinforcement learning (RL), enabling expressive, multimodal policy parameterizations via deterministic probability transport. Compared to diffusion-based policies that rely on stochastic denoising chains, FM uses sampling based on ordinary differential equations (ODEs), with learned velocity fields, which can substantially reduce inference latency and simplify the incorporation of RL objectives. As research in flow-based RL rapidly accelerates across offline continuous control, online fine-tuning, and foundation model alignment, the literature has become highly fragmented. In this survey, we provide a comprehensive taxonomy of flow-matching approaches in reinforcement learning. We organize the literature along two axes: the target distribution being modeled (e.g., action policies, value critics, transition dynamics) and the mechanism of RL signal integration (e.g., energy-weighted regression, flow-based policy gradients, and group relative policy optimization). Furthermore, we survey emerging frontiers such as discrete and non-Euclidean action spaces, provide a systematic comparative analysis against Gaussian and diffusion baselines, and outline critical open problems. Ultimately, this survey serves as a foundational roadmap for the next generation of generative reinforcement learning and alignment.
作者：Nabuat Zaman Nahim, Fairoz Nower Khan, Peizhong Ju
来源
本页事实对应的公开页面，以原文为准。