Learning Translations: Emergent Communication Pretraining for Cooperative Language Acquisition

TL;DR

This paper introduces a new challenge called CLAP that enables agents to learn to communicate across different communities by leveraging pretraining and translation techniques, improving zero-shot coordination.

Contribution

It proposes the CLAP framework and compares two methods, IL and ECTL, for enabling agents to adapt to new communication protocols using prior data.

Findings

ECTL outperforms IL in translating protocols.

Pretraining with EC enhances zero-shot communication.

CLAP facilitates robust cross-community communication.

Abstract

In Emergent Communication (EC) agents learn to communicate with one another, but the protocols that they develop are specialised to their training community. This observation led to research into Zero-Shot Coordination (ZSC) for learning communication strategies that are robust to agents not encountered during training. However, ZSC typically assumes that no prior data is available about the agents that will be encountered in the zero-shot setting. In many cases, this presents an unnecessarily hard problem and rules out communication via preestablished conventions. We propose a novel AI challenge called a Cooperative Language Acquisition Problem (CLAP) in which the ZSC assumptions are relaxed by allowing a 'joiner' agent to learn from a dataset of interactions between agents in a target community. We propose and compare two methods for solving CLAPs: Imitation Learning (IL), and Emergent Communication pretraining and Translation Learning (ECTL), in which an agent is trained in self-play with EC and then learns from the data to translate between the emergent protocol and the target community's protocol.