Beyond Mimicry: Toward Lifelong Adaptability in Imitation Learning

TL;DR

This paper advocates shifting imitation learning focus from memorization to lifelong adaptability, enabling agents to recombine learned behaviors in new contexts without retraining, inspired by cognitive science.

Contribution

It introduces a new research agenda emphasizing compositional adaptability in imitation learning, with metrics, hybrid architectures, and interdisciplinary insights.

Findings

Proposes metrics for compositional generalisation.

Suggests hybrid architectures for adaptable imitation.

Outlines interdisciplinary research directions.

Abstract

Imitation learning stands at a crossroads: despite decades of progress, current imitation learning agents remain sophisticated memorisation machines, excelling at replay but failing when contexts shift or goals evolve. This paper argues that this failure is not technical but foundational: imitation learning has been optimised for the wrong objective. We propose a research agenda that redefines success from perfect replay to compositional adaptability. Such adaptability hinges on learning behavioural primitives once and recombining them through novel contexts without retraining. We establish metrics for compositional generalisation, propose hybrid architectures, and outline interdisciplinary research directions drawing on cognitive science and cultural evolution. Agents that embed adaptability at the core of imitation learning thus have an essential capability for operating in an open-ended world.