Disentangling Dynamics and Content for Control and Planning

TL;DR

This paper proposes a method to learn controllable representations from high-dimensional observations of dynamical systems by leveraging a single informative observation set to enable planning and prediction across multiple observation sets.

Contribution

It introduces a novel approach to disentangle dynamics and content in observations, facilitating control and planning in systems with limited informative data.

Findings

Successfully disentangles dynamics from content in observations.

Enables planning and long-term prediction using learned representations.

Demonstrates effectiveness on high-dimensional dynamical systems.

Abstract

In this paper, We study the problem of learning a controllable representation for high-dimensional observations of dynamical systems. Specifically, we consider a situation where there are multiple sets of observations of dynamical systems with identical underlying dynamics. Only one of these sets has information about the effect of actions on the observation and the rest are just some random observations of the system. Our goal is to utilize the information in that one set and find a representation for the other sets that can be used for planning and ling-term prediction.