Zero-Shot Generalization using Intrinsically Motivated Compositional Emergent Protocols

TL;DR

This paper explores how intrinsically motivated, compositional communication protocols enable artificial agents to generalize to unseen tasks and objects in a zero-shot manner, mimicking human language's compositionality.

Contribution

It introduces a method for developing systematic, unambiguous protolanguages in agents using intrinsic curiosity, facilitating zero-shot generalization and skill transfer.

Findings

Agents can generalize to unseen object interactions.

Skills transfer across tasks without additional training.

Compositional protocols improve zero-shot task performance.

Abstract

Human language has been described as a system that makes \textit{use of finite means to express an unlimited array of thoughts}. Of particular interest is the aspect of compositionality, whereby, the meaning of a compound language expression can be deduced from the meaning of its constituent parts. If artificial agents can develop compositional communication protocols akin to human language, they can be made to seamlessly generalize to unseen combinations. Studies have recognized the role of curiosity in enabling linguistic development in children. In this paper, we seek to use this intrinsic feedback in inducing a systematic and unambiguous protolanguage. We demonstrate how compositionality can enable agents to not only interact with unseen objects but also transfer skills from one task to another in a zero-shot setting: \textit{Can an agent, trained to `pull' and `push twice', `pull twice'?}.