Continual Learning as Computationally Constrained Reinforcement Learning

TL;DR

This paper formalizes continual learning as a framework for developing agents that accumulate knowledge over time, aiming to enhance AI capabilities through long-term skill development.

Contribution

It introduces a formal framework and set of tools for continual learning, providing a foundation for future research in this area.

Findings

Provides a formal definition of continual learning

Introduces a new framework and tools for research

Stimulates further exploration in lifelong AI agents

Abstract

An agent that efficiently accumulates knowledge to develop increasingly sophisticated skills over a long lifetime could advance the frontier of artificial intelligence capabilities. The design of such agents, which remains a long-standing challenge of artificial intelligence, is addressed by the subject of continual learning. This monograph clarifies and formalizes concepts of continual learning, introducing a framework and set of tools to stimulate further research.