Learning Natural Language Inference using Bidirectional LSTM model and Inner-Attention

TL;DR

This paper introduces an improved sentence encoding model for natural language inference that employs a novel inner-attention mechanism, outperforming previous methods on the SNLI dataset with fewer parameters.

Contribution

The paper proposes a new inner-attention mechanism within a bidirectional LSTM framework for sentence encoding in natural language inference tasks.

Findings

Inner-attention improves sentence representations.

Model outperforms existing approaches on SNLI.

Fewer parameters needed for better accuracy.

Abstract

In this paper, we proposed a sentence encoding-based model for recognizing text entailment. In our approach, the encoding of sentence is a two-stage process. Firstly, average pooling was used over word-level bidirectional LSTM (biLSTM) to generate a first-stage sentence representation. Secondly, attention mechanism was employed to replace average pooling on the same sentence for better representations. Instead of using target sentence to attend words in source sentence, we utilized the sentence's first-stage representation to attend words appeared in itself, which is called "Inner-Attention" in our paper . Experiments conducted on Stanford Natural Language Inference (SNLI) Corpus has proved the effectiveness of "Inner-Attention" mechanism. With less number of parameters, our model outperformed the existing best sentence encoding-based approach by a large margin.