Mirac Suzgun


pdf bib
On Evaluating the Generalization of LSTM Models in Formal Languages
Mirac Suzgun | Yonatan Belinkov | Stuart M. Shieber
Proceedings of the Society for Computation in Linguistics (SCiL) 2019

pdf bib
LSTM Networks Can Perform Dynamic Counting
Mirac Suzgun | Yonatan Belinkov | Stuart Shieber | Sebastian Gehrmann
Proceedings of the Workshop on Deep Learning and Formal Languages: Building Bridges

In this paper, we systematically assess the ability of standard recurrent networks to perform dynamic counting and to encode hierarchical representations. All the neural models in our experiments are designed to be small-sized networks both to prevent them from memorizing the training sets and to visualize and interpret their behaviour at test time. Our results demonstrate that the Long Short-Term Memory (LSTM) networks can learn to recognize the well-balanced parenthesis language (Dyck-1) and the shuffles of multiple Dyck-1 languages, each defined over different parenthesis-pairs, by emulating simple real-time k-counter machines. To the best of our knowledge, this work is the first study to introduce the shuffle languages to analyze the computational power of neural networks. We also show that a single-layer LSTM with only one hidden unit is practically sufficient for recognizing the Dyck-1 language. However, none of our recurrent networks was able to yield a good performance on the Dyck-2 language learning task, which requires a model to have a stack-like mechanism for recognition.