English

Equivariant Action Sampling for Reinforcement Learning and Planning

Robotics 2024-12-18 v1 Artificial Intelligence Machine Learning

Abstract

Reinforcement learning (RL) algorithms for continuous control tasks require accurate sampling-based action selection. Many tasks, such as robotic manipulation, contain inherent problem symmetries. However, correctly incorporating symmetry into sampling-based approaches remains a challenge. This work addresses the challenge of preserving symmetry in sampling-based planning and control, a key component for enhancing decision-making efficiency in RL. We introduce an action sampling approach that enforces the desired symmetry. We apply our proposed method to a coordinate regression problem and show that the symmetry aware sampling method drastically outperforms the naive sampling approach. We furthermore develop a general framework for sampling-based model-based planning with Model Predictive Path Integral (MPPI). We compare our MPPI approach with standard sampling methods on several continuous control tasks. Empirical demonstrations across multiple continuous control environments validate the effectiveness of our approach, showcasing the importance of symmetry preservation in sampling-based action selection.

Keywords

Cite

@article{arxiv.2412.12237,
  title  = {Equivariant Action Sampling for Reinforcement Learning and Planning},
  author = {Linfeng Zhao and Owen Howell and Xupeng Zhu and Jung Yeon Park and Zhewen Zhang and Robin Walters and Lawson L. S. Wong},
  journal= {arXiv preprint arXiv:2412.12237},
  year   = {2024}
}

Comments

Published at International Workshop on the Algorithmic Foundations of Robotics (WAFR) 2024. Website: http://lfzhao.com/EquivSampling

R2 v1 2026-06-28T20:37:46.863Z