On the Correspondence between Compositionality and Imitation in Emergent Neural Communication
This work addresses how imitation learning influences language emergence in AI agents, with incremental insights into neural communication dynamics.
The paper investigates the relationship between compositionality and imitation in emergent neural communication, finding that reinforcement learning promotes more compositional languages and that compositional languages are easier to imitate.
Compositionality is a hallmark of human language that not only enables linguistic generalization, but also potentially facilitates acquisition. When simulating language emergence with neural networks, compositionality has been shown to improve communication performance; however, its impact on imitation learning has yet to be investigated. Our work explores the link between compositionality and imitation in a Lewis game played by deep neural agents. Our contributions are twofold: first, we show that the learning algorithm used to imitate is crucial: supervised learning tends to produce more average languages, while reinforcement learning introduces a selection pressure toward more compositional languages. Second, our study reveals that compositional languages are easier to imitate, which may induce the pressure toward compositional languages in RL imitation settings.