How Hard is this Test Set? NLI Characterization by Exploiting Training Dynamics

Adrian Cosma, Stefan Ruseti, Mihai Dascalu, Cornelia Caragea


Abstract
Natural Language Inference (NLI) evaluation is crucial for assessing language understanding models; however, popular datasets suffer from systematic spurious correlations that artificially inflate actual model performance. To address this, we propose a method for the automated creation of a challenging test set without relying on the manual construction of artificial and unrealistic examples. We categorize the test set of popular NLI datasets into three difficulty levels by leveraging methods that exploit training dynamics. This categorization significantly reduces spurious correlation measures, with examples labeled as having the highest difficulty showing markedly decreased performance and encompassing more realistic and diverse linguistic phenomena. When our characterization method is applied to the training set, models trained with only a fraction of the data achieve comparable performance to those trained on the full dataset, surpassing other dataset characterization techniques. Our research addresses limitations in NLI dataset construction, providing a more authentic evaluation of model performance with implications for diverse NLU applications.
Anthology ID:
2024.emnlp-main.175
Volume:
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
Month:
November
Year:
2024
Address:
Miami, Florida, USA
Editors:
Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2990–3001
Language:
URL:
https://aclanthology.org/2024.emnlp-main.175
DOI:
Bibkey:
Cite (ACL):
Adrian Cosma, Stefan Ruseti, Mihai Dascalu, and Cornelia Caragea. 2024. How Hard is this Test Set? NLI Characterization by Exploiting Training Dynamics. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 2990–3001, Miami, Florida, USA. Association for Computational Linguistics.
Cite (Informal):
How Hard is this Test Set? NLI Characterization by Exploiting Training Dynamics (Cosma et al., EMNLP 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.emnlp-main.175.pdf