Unicorn: Continual Learning with a Universal, Off-policy Agent

TL;DR

Unicorn is a novel off-policy agent architecture designed for continual learning in complex, task-agnostic environments, demonstrating superior performance by jointly learning multiple policies in a challenging 3D domain.

Contribution

The paper introduces Unicorn, a universal off-policy agent that effectively handles continual learning without explicit task boundaries by jointly representing multiple policies.

Findings

Unicorn outperforms baseline agents in a complex 3D continual learning domain.

The agent efficiently learns multiple policies simultaneously.

Unicorn demonstrates strong continual learning capabilities in implicit task sequences.

Abstract

Some real-world domains are best characterized as a single task, but for others this perspective is limiting. Instead, some tasks continually grow in complexity, in tandem with the agent's competence. In continual learning, also referred to as lifelong learning, there are no explicit task boundaries or curricula. As learning agents have become more powerful, continual learning remains one of the frontiers that has resisted quick progress. To test continual learning capabilities we consider a challenging 3D domain with an implicit sequence of tasks and sparse rewards. We propose a novel agent architecture called Unicorn, which demonstrates strong continual learning and outperforms several baseline agents on the proposed domain. The agent achieves this by jointly representing and learning multiple policies efficiently, using a parallel off-policy learning setup.