Improving Natural Language Inference with a Pretrained Parser
Deric Pang, Lucy H. Lin, Noah A. Smith
2019-09-18Natural Language Inference
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.
Related Papers
LRCTI: A Large Language Model-Based Framework for Multi-Step Evidence Retrieval and Reasoning in Cyber Threat Intelligence Credibility Verification2025-07-15DS@GT at CheckThat! 2025: Evaluating Context and Tokenization Strategies for Numerical Fact Verification2025-07-08ARAG: Agentic Retrieval Augmented Generation for Personalized Recommendation2025-06-27Thunder-NUBench: A Benchmark for LLMs' Sentence-Level Negation Understanding2025-06-17When Does Meaning Backfire? Investigating the Role of AMRs in NLI2025-06-17Explainable Compliance Detection with Multi-Hop Natural Language Inference on Assurance Case Structure2025-06-10Theorem-of-Thought: A Multi-Agent Framework for Abductive, Deductive, and Inductive Reasoning in Language Models2025-06-08A MISMATCHED Benchmark for Scientific Natural Language Inference2025-06-05