Natural Language Inference over Interaction Space

TL;DR

This paper introduces the Interactive Inference Network (IIN), a neural architecture that leverages interaction tensors to improve understanding in natural language inference, achieving state-of-the-art results on multiple datasets.

Contribution

The paper proposes a novel neural network architecture, DIIN, that utilizes dense interaction tensors for enhanced semantic feature extraction in NLI tasks.

Findings

DIIN achieves state-of-the-art performance on large-scale NLI datasets.

DIIN reduces error rates by over 20% on the MultiNLI dataset.

Interaction tensors contain rich semantic information crucial for NLI.

Abstract

Natural Language Inference (NLI) task requires an agent to determine the logical relationship between a natural language premise and a natural language hypothesis. We introduce Interactive Inference Network (IIN), a novel class of neural network architectures that is able to achieve high-level understanding of the sentence pair by hierarchically extracting semantic features from interaction space. We show that an interaction tensor (attention weight) contains semantic information to solve natural language inference, and a denser interaction tensor contains richer semantic information. One instance of such architecture, Densely Interactive Inference Network (DIIN), demonstrates the state-of-the-art performance on large scale NLI copora and large-scale NLI alike corpus. It's noteworthy that DIIN achieve a greater than 20% error reduction on the challenging Multi-Genre NLI (MultiNLI) dataset with respect to the strongest published system.