Finding Dataset Shortcuts with Grammar Induction

Dan Friedman, Alexander Wettig, Danqi Chen


Abstract
Many NLP datasets have been found to contain shortcuts: simple decision rules that achieve surprisingly high accuracy. However, it is difficult to discover shortcuts automatically. Prior work on automatic shortcut detection has focused on enumerating features like unigrams or bigrams, which can find only low-level shortcuts, or relied on post-hoc model interpretability methods like saliency maps, which reveal qualitative patterns without a clear statistical interpretation. In this work, we propose to use probabilistic grammars to characterize and discover shortcuts in NLP datasets. Specifically, we use a context-free grammar to model patterns in sentence classification datasets and use a synchronous context-free grammar to model datasets involving sentence pairs. The resulting grammars reveal interesting shortcut features in a number of datasets, including both simple and high-level features, and automatically identify groups of test examples on which conventional classifiers fail. Finally, we show that the features we discover can be used to generate diagnostic contrast examples and incorporated into standard robust optimization methods to improve worst-group accuracy.
Anthology ID:
2022.emnlp-main.293
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
4345–4363
Language:
URL:
https://aclanthology.org/2022.emnlp-main.293
DOI:
10.18653/v1/2022.emnlp-main.293
Bibkey:
Cite (ACL):
Dan Friedman, Alexander Wettig, and Danqi Chen. 2022. Finding Dataset Shortcuts with Grammar Induction. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 4345–4363, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Finding Dataset Shortcuts with Grammar Induction (Friedman et al., EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.293.pdf