Multi-Agent Language Learning: Symbolic Mapping

Yicheng Feng, Zongqing Lu

Keywords: emergent communication, multi-agent reinforcement learning

Abstract: The study of emergent communication has long been devoted to coax neural network agents to learn a language sharing similar properties with human language. In this paper, we try to find a natural way to help agents learn a compositional and symmetric language in complex settings like dialog games. Inspired by the theory that human language was originated from simple interactions, we hypothesize that language may evolve from simple tasks to difficult tasks. We propose a novel architecture called symbolic mapping as a basic component of the communication system of agent. We find that symbolic mapping learned in simple referential games can notably promote language learning in difficult tasks. Further, we explore vocabulary expansion, and show that with the help of symbolic mapping, agents can easily learn to use new symbols when the environment becomes more complex. All in all, we probe into how symbolic mapping helps language learning and find that a process from simplicity to complexity can serve as a natural way to help multi-agent language learning.