Bootstrapping Intrinsically Motivated Learning with Human Demonstrations

TL;DR

This paper introduces SGIM-D, an algorithm that combines intrinsic motivation and social learning through demonstrations, enabling efficient learning in complex, continuous environments by acquiring diverse skills and specializing in subspaces.

Contribution

The paper proposes SGIM-D, a novel algorithm that integrates intrinsic motivation with social demonstrations for improved learning in unbounded environments.

Findings

SGIM-D effectively combines social learning and intrinsic motivation.

The algorithm learns a wide repertoire of skills.

SGIM-D specializes in specific subspaces for efficient learning.

Abstract

This paper studies the coupling of internally guided learning and social interaction, and more specifically the improvement owing to demonstrations of the learning by intrinsic motivation. We present Socially Guided Intrinsic Motivation by Demonstration (SGIM-D), an algorithm for learning in continuous, unbounded and non-preset environments. After introducing social learning and intrinsic motivation, we describe the design of our algorithm, before showing through a fishing experiment that SGIM-D efficiently combines the advantages of social learning and intrinsic motivation to gain a wide repertoire while being specialised in specific subspaces.