English

Model Based Planning with Energy Based Models

Machine Learning 2021-03-09 v2 Robotics Machine Learning

Abstract

Model-based planning holds great promise for improving both sample efficiency and generalization in reinforcement learning (RL). We show that energy-based models (EBMs) are a promising class of models to use for model-based planning. EBMs naturally support inference of intermediate states given start and goal state distributions. We provide an online algorithm to train EBMs while interacting with the environment, and show that EBMs allow for significantly better online learning than corresponding feed-forward networks. We further show that EBMs support maximum entropy state inference and are able to generate diverse state space plans. We show that inference purely in state space - without planning actions - allows for better generalization to previously unseen obstacles in the environment and prevents the planner from exploiting the dynamics model by applying uncharacteristic action sequences. Finally, we show that online EBM training naturally leads to intentionally planned state exploration which performs significantly better than random exploration.

Keywords

Cite

@article{arxiv.1909.06878,
  title  = {Model Based Planning with Energy Based Models},
  author = {Yilun Du and Toru Lin and Igor Mordatch},
  journal= {arXiv preprint arXiv:1909.06878},
  year   = {2021}
}

Comments

CoRL 2019

R2 v1 2026-06-23T11:15:52.522Z