How can self-attention networks recognize Dyck-n languages?

Abstract

We focus on the recognition of Dyck-n (Dn) languages with self-attention (SA) networks, which has been deemed to be a difficult task for these networks. We compare the performance of two variants of SA, one with a starting symbol (SA+) and one without (SA−). Our results show that SA+ is able to generalize to longer sequences and deeper dependencies. For D2, we find that SA− completely breaks down on long sequences whereas the accuracy of SA+ is 58.82%. We find attention maps learned by SA+ to be amenable to interpretation and compatible with a stack-based language recognizer. Surprisingly, the performance of SA networks is at par with LSTMs, which provides evidence on the ability of SA to learn hierarchies without recursion.