Model Based Planning with Energy Based Models

By Yilun Du, Toru Lin, Igor Mordatch. 2019

44 citations recorded by the source corpus.

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.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Implicit Behavioral Cloning (2021, neutral): “Other recent work [14] uses EBMs in a model-based planning framework, or 230 uses EBMs in imitation learning [48] but with an on-policy algorithm.”
Improved Contrastive Divergence Training of Energy Based Models (2020, background): “, 2017), reinforcement learning (Haarnoja et al., 2017; Du et al., 2019), continual learning (Li et al.”
Energy-Based Models for Continual Learning (2020, background): “, 2020), reinforcement learning (Haarnoja et al., 2017; Du et al., 2019), image generation (Salakhutdinov & Hinton, 2009; Xie et al.”