Atomic-SNLI: Fine-Grained Natural Language Inference through Atomic Fact Decomposition

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Huang, Minghui
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912814510112768
author Huang, Minghui
author_facet Huang, Minghui
contents Current Natural Language Inference (NLI) systems primarily operate at the sentence level, providing black-box decisions that lack explanatory power. While atomic-level NLI offers a promising alternative by decomposing hypotheses into individual facts, we demonstrate that the conventional assumption that a hypothesis is entailed only when all its atomic facts are entailed fails in practice due to models' poor performance on fine-grained reasoning. Our analysis reveals that existing models perform substantially worse on atomic level inference compared to sentence level tasks. To address this limitation, we introduce Atomic-SNLI, a novel dataset constructed by decomposing SNLI and enriching it with carefully curated atomic level examples through linguistically informed generation strategies. Experimental results demonstrate that models fine-tuned on Atomic-SNLI achieve significant improvements in atomic reasoning capabilities while maintaining strong sentence level performance, enabling both accurate judgements and transparent, explainable results at the fact level.
format Preprint
id arxiv_https___arxiv_org_abs_2601_06528
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Atomic-SNLI: Fine-Grained Natural Language Inference through Atomic Fact Decomposition
Huang, Minghui
Computation and Language
Artificial Intelligence
Current Natural Language Inference (NLI) systems primarily operate at the sentence level, providing black-box decisions that lack explanatory power. While atomic-level NLI offers a promising alternative by decomposing hypotheses into individual facts, we demonstrate that the conventional assumption that a hypothesis is entailed only when all its atomic facts are entailed fails in practice due to models' poor performance on fine-grained reasoning. Our analysis reveals that existing models perform substantially worse on atomic level inference compared to sentence level tasks. To address this limitation, we introduce Atomic-SNLI, a novel dataset constructed by decomposing SNLI and enriching it with carefully curated atomic level examples through linguistically informed generation strategies. Experimental results demonstrate that models fine-tuned on Atomic-SNLI achieve significant improvements in atomic reasoning capabilities while maintaining strong sentence level performance, enabling both accurate judgements and transparent, explainable results at the fact level.
title Atomic-SNLI: Fine-Grained Natural Language Inference through Atomic Fact Decomposition
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2601.06528