Table-Text Alignment: Explaining Claim Verification Against Tables in Scientific Papers

TL;DR

This paper introduces a new approach to scientific claim verification against tables by focusing on explainability through cell-level alignment, enhancing interpretability and revealing limitations of current large language models.

Contribution

The paper presents a novel dataset with human-annotated rationales for claim verification and demonstrates that incorporating alignment improves performance while exposing reasoning gaps in LLMs.

Findings

Alignment-based methods outperform baseline models.

Most LLMs fail to identify human-aligned rationales.

Incorporating rationales enhances claim verification accuracy.

Abstract

Scientific claim verification against tables typically requires predicting whether a claim is supported or refuted given a table. However, we argue that predicting the final label alone is insufficient: it reveals little about the model's reasoning and offers limited interpretability. To address this, we reframe table-text alignment as an explanation task, requiring models to identify the table cells essential for claim verification. We build a new dataset by extending the SciTab benchmark with human-annotated cell-level rationales. Annotators verify the claim label and highlight the minimal set of cells needed to support their decision. After the annotation process, we utilize the collected information and propose a taxonomy for handling ambiguous cases. Our experiments show that (i) incorporating table alignment information improves claim verification performance, and (ii) most LLMs, while often predicting correct labels, fail to recover human-aligned rationales, suggesting that their predictions do not stem from faithful reasoning.