Machine Reading, Fast and Slow: When Do Models “Understand” Language?

Sagnik Ray Choudhury, Anna Rogers, Isabelle Augenstein


Abstract
Two of the most fundamental issues in Natural Language Understanding (NLU) at present are: (a) how it can established whether deep learning-based models score highly on NLU benchmarks for the ”right” reasons; and (b) what those reasons would even be. We investigate the behavior of reading comprehension models with respect to two linguistic ”skills”: coreference resolution and comparison. We propose a definition for the reasoning steps expected from a system that would be ”reading slowly”, and compare that with the behavior of five models of the BERT family of various sizes, observed through saliency scores and counterfactual explanations. We find that for comparison (but not coreference) the systems based on larger encoders are more likely to rely on the ”right” information, but even they struggle with generalization, suggesting that they still learn specific lexical patterns rather than the general principles of comparison.
Anthology ID:
2022.coling-1.8
Volume:
Proceedings of the 29th International Conference on Computational Linguistics
Month:
October
Year:
2022
Address:
Gyeongju, Republic of Korea
Venue:
COLING
SIG:
Publisher:
International Committee on Computational Linguistics
Note:
Pages:
78–93
Language:
URL:
https://aclanthology.org/2022.coling-1.8
DOI:
Bibkey:
Cite (ACL):
Sagnik Ray Choudhury, Anna Rogers, and Isabelle Augenstein. 2022. Machine Reading, Fast and Slow: When Do Models “Understand” Language?. In Proceedings of the 29th International Conference on Computational Linguistics, pages 78–93, Gyeongju, Republic of Korea. International Committee on Computational Linguistics.
Cite (Informal):
Machine Reading, Fast and Slow: When Do Models “Understand” Language? (Ray Choudhury et al., COLING 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.coling-1.8.pdf
Data
HotpotQAQEDQuoref