paper-with-me

홈 › Papers

Iterative Refinement of Flow Policies in Probability Space for Online Reinforcement Learning

2025-10-17 · Mingyang Sun, Pengxiang Ding, Weinan Zhang, Donglin Wang arxiv

While behavior cloning with flow/diffusion policies excels at learning complex skills from demonstrations, it remains vulnerable to distributional shift, and standard RL methods struggle to fine-tune these models due to their iterative inference process and the limitations of existing workarounds. In this work, we introduce the Stepwise Flow Policy (SWFP) framework, founded on the key insight that discretizing the flow matching inference process via a fixed-step Euler scheme inherently aligns it with the variational Jordan-Kinderlehrer-Otto (JKO) principle from optimal transport. SWFP decomposes the global flow into a sequence of small, incremental transformations between proximate distributions. Each step corresponds to a JKO update, regularizing policy changes to stay near the previous iterate and ensuring stable online adaptation with entropic regularization. This decomposition yields an efficient algorithm that fine-tunes pre-trained flows via a cascade of small flow blocks, offering significant advantages: simpler/faster training of sub-models, reduced computational/memory costs, and provable stability grounded in Wasserstein trust regions. Comprehensive experiments demonstrate SWFP's enhanced stability, efficiency, and superior adaptation performance across diverse robotic control benchmarks.

📄 PDF Abstract BibTeX arXiv:2510.15388

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

SUREFlow: State-space Uncertainty-aware REsidual Flow Matching for Robust Robot Manipulation

2026-07-11 · Md Tanvir Islam, Sai Navaneet Peddapalli, Sangmoon Lee, Sangtae Ahn arxiv

Generative vision-language-action policies have advanced robot manipulation, but they often exhibit instability under noise, partial observability, and stochastic initial conditions. During extended rollouts, small veloc…

Computational EfficiencyRobot Manipulation

Discrete MeanFlow: One-Step Generation via Conditional Transition Kernels

2026-05-12 · Fairoz Nower Khan, Nabuat Zaman Nahim, Md Sajid Ahmed, Ruiquan Huang 외 arxiv

MeanFlow enables one-step generation in continuous spaces by learning an average velocity over a time interval rather than the instantaneous velocity field of flow matching. However, discrete state spaces do not have smo…

DeFlow: Decoupling Manifold Modeling and Value Maximization for Offline Policy Extraction

2026-01-15 · Zhancun Mu arxiv

We present DeFlow, a decoupled offline RL framework that leverages flow matching to faithfully capture complex behavior manifolds. Optimizing generative policies is computationally prohibitive, typically necessitating ba…

Offline RL

TGPR: Tree-Guided Policy Refinement for Robust Self-Debugging of LLMs

2025-10-08 · Daria Ozerova, Ekaterina Trofimova arxiv

Iterative refinement has been a promising paradigm to enable large language models (LLMs) to resolve difficult reasoning and problem-solving tasks. One of the key challenges, however, is how to effectively search through…

Aligning Flow Map Policies with Optimal Q-Guidance

2026-05-12 · Christos Ziakas, Alessandra Russo, Avishek Joey Bose arxiv

Generative policies based on expressive model classes, such as diffusion and flow matching, are well-suited to complex control problems with highly multimodal action distributions. Their expressivity, however, comes at a…

Reinforcement Learning