Self-Explaining Deviations for Coordination

TL;DR

This paper introduces a novel algorithm, IMPROVISED, enabling agents in multi-agent coordination tasks to perform self-explaining deviations that reveal abnormal circumstances, enhancing understanding and strategic play.

Contribution

The paper formalizes self-explaining deviations and presents IMPROVISED, the first method to generate finesse plays in Hanabi, demonstrating improved coordination and theory of mind capabilities.

Findings

IMPROVISED successfully performs self-explaining deviations in toy and Hanabi settings.

It is the first method to produce finesse plays in Hanabi.

Results show enhanced agent coordination and interpretability.

Abstract

Fully cooperative, partially observable multi-agent problems are ubiquitous in the real world. In this paper, we focus on a specific subclass of coordination problems in which humans are able to discover self-explaining deviations (SEDs). SEDs are actions that deviate from the common understanding of what reasonable behavior would be in normal circumstances. They are taken with the intention of causing another agent or other agents to realize, using theory of mind, that the circumstance must be abnormal. We first motivate SED with a real world example and formalize its definition. Next, we introduce a novel algorithm, improvement maximizing self-explaining deviations (IMPROVISED), to perform SEDs. Lastly, we evaluate IMPROVISED both in an illustrative toy setting and the popular benchmark setting Hanabi, where it is the first method to produce so called finesse plays, which are regarded as one of the more iconic examples of human theory of mind.