Latent Skill Planning for Exploration and Transfer

TL;DR

This paper presents a reinforcement learning framework that combines learned skills with world models through partial amortization, enabling fast adaptation and transfer across complex tasks with improved sample efficiency.

Contribution

It introduces a novel integration of skills and world models using online planning and partial amortization for enhanced transfer and adaptation in RL agents.

Findings

Improved sample efficiency in locomotion tasks.

Effective transfer learning across tasks.

Enhanced adaptation speed compared to baselines.

Abstract

To quickly solve new tasks in complex environments, intelligent agents need to build up reusable knowledge. For example, a learned world model captures knowledge about the environment that applies to new tasks. Similarly, skills capture general behaviors that can apply to new tasks. In this paper, we investigate how these two approaches can be integrated into a single reinforcement learning agent. Specifically, we leverage the idea of partial amortization for fast adaptation at test time. For this, actions are produced by a policy that is learned over time while the skills it conditions on are chosen using online planning. We demonstrate the benefits of our design decisions across a suite of challenging locomotion tasks and demonstrate improved sample efficiency in single tasks as well as in transfer from one task to another, as compared to competitive baselines. Videos are available at: https://sites.google.com/view/latent-skill-planning/