Timezone: »

 
Poster
Symmetry-Based Disentangled Representation Learning requires Interaction with Environments
Hugo Caselles-Dupré · Michael Garcia Ortiz · David Filliat

Tue Dec 10 05:30 PM -- 07:30 PM (PST) @ East Exhibition Hall B + C #64
in Algorithms -- Unsupervised Learning »

Finding a generally accepted formal definition of a disentangled representation in the context of an agent behaving in an environment is an important challenge towards the construction of data-efficient autonomous agents. Higgins et al. recently proposed Symmetry-Based Disentangled Representation Learning, a definition based on a characterization of symmetries in the environment using group theory. We build on their work and make observations, theoretical and empirical, that lead us to argue that Symmetry-Based Disentangled Representation Learning cannot only be based on static observations: agents should interact with the environment to discover its symmetries. Our experiments can be reproduced in Colab and the code is available on GitHub.

Keywords: Representation Learning · Reinforcement Learning · Algorithms · Deep Learning -> Predictive Models; Reinforcement Learning and Planning

Author Information

Hugo Caselles-Dupré (Flowers Laboratory (ENSTA ParisTech & INRIA) & Softbank Robotics Europe)

Postdoc working on Reinforcement Learning and Developmental Robotics.

Michael Garcia Ortiz (SoftBank Robotics Europe)
David Filliat (ENSTA)

More from the Same Authors