Diffusion Policy with Behavioral Advantage Correction for Offline Reinforcement Learning
2026-08-03 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors address a problem in offline reinforcement learning where the learned policy’s value estimates can be wrong because it differs from the original behavior data. They create a method called BAC-PE that uses the behavior policy’s value function to correct these estimates, reducing mistakes like overestimating values. They also use diffusion models to better match the learned policy to the behavior policy and guide improvements with corrected values. Their combined approach, DPBAC, performs better than existing methods on standard testing tasks. The authors back their work with theoretical analysis and experiments.
Offline Reinforcement LearningQ-value estimationBehavior policyPolicy evaluationDistribution shiftDiffusion modelsPolicy optimizationPessimistic conservatismBehavioral advantageD4RL benchmark
Authors
Botao Dong, Longyang Huang, Ning Pang, Hongtian Chen
Abstract
In offline reinforcement learning (RL), the distribution shift between behavioral data and the learned policy can lead to erroneous \emph{Q}-value estimation, thereby misguiding the direction of policy optimization. To address this issue, we develop a behavioral advantage corrected policy evaluation (BAC-PE) approach, which utilizes the \emph{Q}-function of the behavior policy to correct the learned policy's \emph{Q}-function, thus mitigating pessimistic conservatism and overestimation bias. Furthermore, the convergence of BAC-PE is analyzed theoretically, and an upper bound on the difference between the learned \emph{Q}-function and the true \emph{Q}-function is derived. To alleviate distribution shift, this work employs diffusion models to represent both the behavior policy and the learned policy, performing distribution matching for accurate policy regularization. Additionally, \emph{Q}-value guidance is incorporated into the training process to achieve effective policy improvement. By combining BAC-PE with diffusion policy modeling, we propose the diffusion policy with behavioral advantage correction (DPBAC) algorithm. Compared to existing offline methods, DPBAC demonstrates stronger policy representation capabilities and effectively mitigates the bias in \emph{Q}-value estimation. Experimental results on multiple domains of D4RL tasks show that DPBAC achieves superior performance, with notable advantages over state-of-the-art (SOTA) algorithms.