Catalytic Role of Noise and Necessity of Inductive Biases in the Emergence of Compositional Communication

 We will talk about our recent compositionality paper accepted at NeurIPS 2021. Communication is compositional if complex signals can be represented as a combi-nation of simpler subparts. In this paper, we theoretically show that inductive biases on both the training framework and the data are needed to develop a compositional communication. Moreover, we prove that compositionality spontaneously arises in the signaling games, where agents communicate over a noisy channel. We experi-mentally confirm that a range of noise levels, which depends on the model and the data, indeed promotes compositionality. Finally, we provide a comprehensive study of this dependence and report results in terms of recently studied compositionality metrics: topographical similarity, conflict count, and context independence