Character-level Intra Attention Network for Natural Language Inference

TL;DR

This paper introduces CIAN, a character-level neural network with intra-attention for natural language inference, achieving improved performance on the MNLI dataset by capturing intra-sentence semantics more effectively.

Contribution

The paper presents a novel character-level convolutional network combined with intra-attention for NLI, replacing traditional word embeddings and enhancing sentence understanding.

Findings

Improved accuracy on MNLI dataset

Effective intra-sentence semantic capture

Outperforms previous models on NLI task

Abstract

Natural language inference (NLI) is a central problem in language understanding. End-to-end artificial neural networks have reached state-of-the-art performance in NLI field recently. In this paper, we propose Character-level Intra Attention Network (CIAN) for the NLI task. In our model, we use the character-level convolutional network to replace the standard word embedding layer, and we use the intra attention to capture the intra-sentence semantics. The proposed CIAN model provides improved results based on a newly published MNLI corpus.