English

Improving Natural Language Inference with a Pretrained Parser

Computation and Language 2019-09-19 v1

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.

Keywords

Cite

@article{arxiv.1909.08217,
  title  = {Improving Natural Language Inference with a Pretrained Parser},
  author = {Deric Pang and Lucy H. Lin and Noah A. Smith},
  journal= {arXiv preprint arXiv:1909.08217},
  year   = {2019}
}
R2 v1 2026-06-23T11:18:46.546Z