The Irrationality of Neural Rationale Models

Neural rationale models are popular for interpretable predictions of NLP tasks. In these, a selector extracts segments of the input text, called rationales, and passes these segments to a classifier for prediction. Since the rationale is the only information accessible to the classifier, it is plaus...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Zheng, Yiming, Booth, Serena, Shah, Julie, Zhou, Yilun
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 24.07.2022
Subjects
Classifiers
Empirical analysis
Irrationality
Segments
Online AccessGet full text

Cover

More Information
Summary:Neural rationale models are popular for interpretable predictions of NLP tasks. In these, a selector extracts segments of the input text, called rationales, and passes these segments to a classifier for prediction. Since the rationale is the only information accessible to the classifier, it is plausibly defined as the explanation. Is such a characterization unconditionally correct? In this paper, we argue to the contrary, with both philosophical perspectives and empirical evidence suggesting that rationale models are, perhaps, less rational and interpretable than expected. We call for more rigorous and comprehensive evaluations of these models to ensure desired properties of interpretability are indeed achieved. The code can be found at https://github.com/yimingz89/Neural-Rationale-Analysis.
ISSN:2331-8422