Using Regular Languages to Explore the Representational Capacity of Recurrent Neural Architectures

TL;DR

This paper investigates how well various recurrent neural network architectures can handle long-distance dependencies in sequences by using synthetic datasets based on subregular languages with controllable complexity.

Contribution

It introduces a benchmarking approach using synthetic SPk datasets to systematically evaluate RNNs' capacity to model increasing levels of long-distance dependencies.

Findings

LDDs significantly impact RNN performance

Synthetic datasets effectively benchmark RNNs' capabilities

SPk languages serve as useful controlled test cases

Abstract

The presence of Long Distance Dependencies (LDDs) in sequential data poses significant challenges for computational models. Various recurrent neural architectures have been designed to mitigate this issue. In order to test these state-of-the-art architectures, there is growing need for rich benchmarking datasets. However, one of the drawbacks of existing datasets is the lack of experimental control with regards to the presence and/or degree of LDDs. This lack of control limits the analysis of model performance in relation to the specific challenge posed by LDDs. One way to address this is to use synthetic data having the properties of subregular languages. The degree of LDDs within the generated data can be controlled through the k parameter, length of the generated strings, and by choosing appropriate forbidden strings. In this paper, we explore the capacity of different RNN extensions to model LDDs, by evaluating these models on a sequence of SPk synthesized datasets, where each subsequent dataset exhibits a longer degree of LDD. Even though SPk are simple languages, the presence of LDDs does have significant impact on the performance of recurrent neural architectures, thus making them prime candidate in benchmarking tasks.