Emergent Communication

Emergent communication research investigates how language-like communication systems arise spontaneously in multi-agent systems, primarily using deep reinforcement learning models and variations of the Lewis signaling game. Current research focuses on improving the compositionality, interpretability, and efficiency of emergent languages, often by incorporating attention mechanisms, inductive biases, and information bottleneck principles into model architectures. This field offers valuable insights into the origins and structure of human language, and has potential applications in areas such as human-computer interaction, multi-agent robotics, and network optimization.