Saved in:
Bibliographic Details
Main Authors: Celik, Onur, Li, Zechu, Blessing, Denis, Li, Ge, Palenicek, Daniel, Peters, Jan, Chalvatzaki, Georgia, Neumann, Gerhard
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.02316
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909644772868096
author Celik, Onur
Li, Zechu
Blessing, Denis
Li, Ge
Palenicek, Daniel
Peters, Jan
Chalvatzaki, Georgia
Neumann, Gerhard
author_facet Celik, Onur
Li, Zechu
Blessing, Denis
Li, Ge
Palenicek, Daniel
Peters, Jan
Chalvatzaki, Georgia
Neumann, Gerhard
contents Maximum entropy reinforcement learning (MaxEnt-RL) has become the standard approach to RL due to its beneficial exploration properties. Traditionally, policies are parameterized using Gaussian distributions, which significantly limits their representational capacity. Diffusion-based policies offer a more expressive alternative, yet integrating them into MaxEnt-RL poses challenges-primarily due to the intractability of computing their marginal entropy. To overcome this, we propose Diffusion-Based Maximum Entropy RL (DIME). \emph{DIME} leverages recent advances in approximate inference with diffusion models to derive a lower bound on the maximum entropy objective. Additionally, we propose a policy iteration scheme that provably converges to the optimal diffusion policy. Our method enables the use of expressive diffusion-based policies while retaining the principled exploration benefits of MaxEnt-RL, significantly outperforming other diffusion-based methods on challenging high-dimensional control benchmarks. It is also competitive with state-of-the-art non-diffusion based RL methods while requiring fewer algorithmic design choices and smaller update-to-data ratios, reducing computational complexity.
format Preprint
id arxiv_https___arxiv_org_abs_2502_02316
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DIME:Diffusion-Based Maximum Entropy Reinforcement Learning
Celik, Onur
Li, Zechu
Blessing, Denis
Li, Ge
Palenicek, Daniel
Peters, Jan
Chalvatzaki, Georgia
Neumann, Gerhard
Machine Learning
Maximum entropy reinforcement learning (MaxEnt-RL) has become the standard approach to RL due to its beneficial exploration properties. Traditionally, policies are parameterized using Gaussian distributions, which significantly limits their representational capacity. Diffusion-based policies offer a more expressive alternative, yet integrating them into MaxEnt-RL poses challenges-primarily due to the intractability of computing their marginal entropy. To overcome this, we propose Diffusion-Based Maximum Entropy RL (DIME). \emph{DIME} leverages recent advances in approximate inference with diffusion models to derive a lower bound on the maximum entropy objective. Additionally, we propose a policy iteration scheme that provably converges to the optimal diffusion policy. Our method enables the use of expressive diffusion-based policies while retaining the principled exploration benefits of MaxEnt-RL, significantly outperforming other diffusion-based methods on challenging high-dimensional control benchmarks. It is also competitive with state-of-the-art non-diffusion based RL methods while requiring fewer algorithmic design choices and smaller update-to-data ratios, reducing computational complexity.
title DIME:Diffusion-Based Maximum Entropy Reinforcement Learning
topic Machine Learning
url https://arxiv.org/abs/2502.02316