Capacity, Bandwidth, and Compositionality in Emergent Language Learning
This addresses the problem of understanding learning biases in emergent languages for researchers in AI and linguistics, but it is incremental as it builds on prior work focusing on bandwidth.
The paper investigates how neural network capacity and channel bandwidth affect the learning of compositional emergent languages, finding evidence for a lower bound but not an upper bound on capacity that induces compositionality.
Many recent works have discussed the propensity, or lack thereof, for emergent languages to exhibit properties of natural languages. A favorite in the literature is learning compositionality. We note that most of those works have focused on communicative bandwidth as being of primary importance. While important, it is not the only contributing factor. In this paper, we investigate the learning biases that affect the efficacy and compositionality of emergent languages. Our foremost contribution is to explore how capacity of a neural network impacts its ability to learn a compositional language. We additionally introduce a set of evaluation metrics with which we analyze the learned languages. Our hypothesis is that there should be a specific range of model capacity and channel bandwidth that induces compositional structure in the resulting language and consequently encourages systematic generalization. While we empirically see evidence for the bottom of this range, we curiously do not find evidence for the top part of the range and believe that this is an open question for the community.