MorphNLI: A Stepwise Approach to Natural Language Inference Using Text Morphing

TL;DR

MorphNLI introduces a modular, stepwise method for natural language inference that transforms premises into hypotheses through atomic edits, improving accuracy especially across domains and enhancing interpretability.

Contribution

The paper presents a novel stepwise approach to NLI using text morphing and atomic edits, which improves cross-domain performance and provides explainability.

Findings

Outperforms strong baselines by up to 12.6% in cross-domain settings

Enhances interpretability through atomic edit analysis

Demonstrates robustness across diverse NLI datasets

Abstract

We introduce MorphNLI, a modular step-by-step approach to natural language inference (NLI). When classifying the premise-hypothesis pairs into {entailment, contradiction, neutral}, we use a language model to generate the necessary edits to incrementally transform (i.e., morph) the premise into the hypothesis. Then, using an off-the-shelf NLI model we track how the entailment progresses with these atomic changes, aggregating these intermediate labels into a final output. We demonstrate the advantages of our proposed method particularly in realistic cross-domain settings, where our method always outperforms strong baselines with improvements up to 12.6% (relative). Further, our proposed approach is explainable as the atomic edits can be used to understand the overall NLI label.