Reinforcement Learning with Prototypical RepresentationsDownload PDF

Published: 15 Jun 2022, Last Modified: 22 Oct 2023SSL-RL 2021 SpotlightReaders: Everyone
Keywords: RL, SSL, Unsupervised RL, Image-Based RL
TL;DR: Framework for task-agnostic exploration and representation learning in image-based RL
Abstract: Learning effective representations in image-based environments is crucial for sample efficient Reinforcement Learning (RL). Unfortunately, in RL, representation learning is confounded with the exploratory experience of the agent -- learning a useful representation requires diverse data, while effective exploration is only possible with coherent representations. Furthermore, we would like to learn representations that not only generalize across tasks but also accelerate downstream exploration for efficient task-specific training. To address these challenges we propose Proto-RL, a self-supervised framework that ties representation learning with exploration through prototypical representations. These prototypes simultaneously serve as a summarization of the exploratory experience of an agent as well as a basis for representing observations. We pre-train these task-agnostic representations and prototypes on environments without downstream task information. This enables state-of-the-art downstream RL on a set of difficult continuous control tasks.
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 2 code implementations](https://www.catalyzex.com/paper/arxiv:2102.11271/code)
0 Replies

Loading