Structured Attention Networks

TL;DR

This paper introduces structured attention networks that incorporate graphical model-based structural dependencies into deep neural networks, enhancing their ability to model complex structures in tasks like translation and question answering.

Contribution

It presents a simple extension of attention mechanisms to include structured distributions, enabling richer dependency modeling within end-to-end trainable neural networks.

Findings

Structured attention outperforms baseline models on multiple tasks.

Models learn meaningful unsupervised hidden representations.

Effective incorporation of structural biases improves task performance.

Abstract

Attention networks have proven to be an effective approach for embedding categorical inference within a deep neural network. However, for many tasks we may want to model richer structural dependencies without abandoning end-to-end training. In this work, we experiment with incorporating richer structural distributions, encoded using graphical models, within deep networks. We show that these structured attention networks are simple extensions of the basic attention procedure, and that they allow for extending attention beyond the standard soft-selection approach, such as attending to partial segmentations or to subtrees. We experiment with two different classes of structured attention networks: a linear-chain conditional random field and a graph-based parsing model, and describe how these models can be practically implemented as neural network layers. Experiments show that this approach is effective for incorporating structural biases, and structured attention networks outperform baseline attention models on a variety of synthetic and real tasks: tree transduction, neural machine translation, question answering, and natural language inference. We further find that models trained in this way learn interesting unsupervised hidden representations that generalize simple attention.