Learning to Request Guidance in Emergent Communication

TL;DR

This paper introduces a bidirectional communication framework where an agent can request guidance from a pre-trained guide during learning, leading to adaptive help-seeking behavior based on uncertainty, which improves learning efficiency.

Contribution

It extends previous unidirectional emergent communication models by enabling the agent to ask for help, with a gating mechanism that reduces guidance requests over time.

Findings

Guidance requests decrease as training progresses.

Requests are made primarily in high-uncertainty situations.

The gating mechanism effectively controls guidance flow.

Abstract

Previous research into agent communication has shown that a pre-trained guide can speed up the learning process of an imitation learning agent. The guide achieves this by providing the agent with discrete messages in an emerged language about how to solve the task. We extend this one-directional communication by a one-bit communication channel from the learner back to the guide: It is able to ask the guide for help, and we limit the guidance by penalizing the learner for these requests. During training, the agent learns to control this gate based on its current observation. We find that the amount of requested guidance decreases over time and guidance is requested in situations of high uncertainty. We investigate the agent's performance in cases of open and closed gates and discuss potential motives for the observed gating behavior.