Decentralised Emergence of Robust and Adaptive Linguistic Conventions in Populations of Autonomous Agents Grounded in Continuous Worlds

This paper introduces a methodology for establishing linguistic conventions in populations of autonomous agents in a fully decentralised manner. As agents take part in local communicative interactions, they gradually establish a common conceptual system and vocabulary that enables them to communicate about arbitrary entities in their continuous environment. Apart from introducing the methodology, we also present six experiments that showcase the robustness of the methodology against sensor defects, its ability to handle noisy observations and uncalibrated sensors, and its suitability for continual learning.