跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2602.06138 · v4

Flow Matching for Offline Reinforcement Learning with Discrete Actions

首次提交 2026-02-06 03:13
版本更新 2026-11-15 14:27
雷达观测 2026-11-17 19:15
arXiv ↗PDF ↗

摘要

Generative policies based on diffusion models and flow matching have shown strong promise for offline reinforcement learning (RL), but their applicability remains largely confined to continuous action spaces. To address a broader range of offline RL settings, we extend flow matching to a general framework that supports discrete action spaces with multiple objectives. Specifically, we replace continuous flows with continuous-time Markov chains, trained using a Q-weighted flow matching objective. We then extend our design to multi-agent settings, mitigating the exponential growth of joint action spaces via a factorized conditional path. For an oracle single policy-improvement step from the behavior distribution, we theoretically show that, under idealized conditions, the Q-weighted objective recovers the KL-regularized Boltzmann policy. Extensive experiments demonstrate that our method performs robustly across diverse settings and benchmarks, including high-dimensional control, multi-agent games, and dynamically changing preferences over multiple objectives, while achieving competitive performance and demonstrating advantages in multimodal decision-making and preference-conditioned policy representation. Our discrete framework can also be applied to continuous-control problems through action quantization, providing a flexible trade-off between representational complexity and performance.

作者备注:Accepted at NeurIPS 2026

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v4,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v4 · 2026-11-15 14:27 · 最新观测版本

Flow Matching for Offline Reinforcement Learning with Discrete Actions

Generative policies based on diffusion models and flow matching have shown strong promise for offline reinforcement learning (RL), but their applicability remains largely confined to continuous action spaces. To address a broader range of offline RL settings, we extend flow matching to a general framework that supports discrete action spaces with multiple objectives. Specifically, we replace continuous flows with continuous-time Markov chains, trained using a Q-weighted flow matching objective. We then extend our design to multi-agent settings, mitigating the exponential growth of joint action spaces via a factorized conditional path. For an oracle single policy-improvement step from the behavior distribution, we theoretically show that, under idealized conditions, the Q-weighted objective recovers the KL-regularized Boltzmann policy. Extensive experiments demonstrate that our method performs robustly across diverse settings and benchmarks, including high-dimensional control, multi-agent games, and dynamically changing preferences over multiple objectives, while achieving competitive performance and demonstrating advantages in multimodal decision-making and preference-conditioned policy representation. Our discrete framework can also be applied to continuous-control problems through action quantization, providing a flexible trade-off between representational complexity and performance.

查看此版本解读 →查看此版本原文 ↗