IPD: Boosting Sequential Policy with Imaginary Planning Distillation in Offline Reinforcement Learning

TL;DR

IPD introduces a novel offline RL framework that enhances sequential policy learning by integrating imaginary planning and distillation, leading to improved decision-making and performance across benchmarks.

Contribution

The paper proposes Imaginary Planning Distillation (IPD), a new method that combines offline planning, world modeling, and policy distillation to overcome limitations of existing offline RL models.

Findings

IPD outperforms state-of-the-art offline RL methods on D4RL benchmarks.

IPD effectively integrates imagined optimal trajectories to improve policy quality.

IPD enhances decision stability and inference performance.

Abstract

Decision transformer based sequential policies have emerged as a powerful paradigm in offline reinforcement learning (RL), yet their efficacy remains constrained by the quality of static datasets and inherent architectural limitations. Specifically, these models often struggle to effectively integrate suboptimal experiences and fail to explicitly plan for an optimal policy. To bridge this gap, we propose \textbf{Imaginary Planning Distillation (IPD)}, a novel framework that seamlessly incorporates offline planning into data generation, supervised training, and online inference. Our framework first learns a world model equipped with uncertainty measures and a quasi-optimal value function from the offline data. These components are utilized to identify suboptimal trajectories and augment them with reliable, imagined optimal rollouts generated via Model Predictive Control (MPC). A Transformer-based sequential policy is then trained on this enriched dataset, complemented by a value-guided objective that promotes the distillation of the optimal policy. By replacing the conventional, manually-tuned return-to-go with the learned quasi-optimal value function, IPD improves both decision-making stability and performance during inference. Empirical evaluations on the D4RL benchmark demonstrate that IPD significantly outperforms several state-of-the-art value-based and transformer-based offline RL methods across diverse tasks.