Improving Natural Language Inference with a Pretrained Parser

Abstract

We introduce a novel approach to incorporate syntax into natural language inference (NLI) models. Our method uses contextual token-level vector representations from a pretrained dependency parser. Like other contextual embedders, our method is broadly applicable to any neural model. We experiment with four strong NLI models (decomposable attention model, ESIM, BERT, and MT-DNN), and show consistent benefit to accuracy across three NLI benchmarks.

References in corpus (1)

• Enhanced LSTM for Natural Language Inference

Cited by in corpus (4)

• Syntactic Data Augmentation Increases Robustness to Inference Heuristics
• The Curse of Performance Instability in Analysis Datasets: Consequences, Source, and Suggestions
• ExBERT: An External Knowledge Enhanced BERT for Natural Language Inference
• The Limitations of Limited Context for Constituency Parsing