Undersensitivity in Neural Reading Comprehension

Johannes Welbl, Pasquale Minervini, Max Bartolo, Pontus Stenetorp, Sebastian Riedel


Abstract
Current reading comprehension methods generalise well to in-distribution test sets, yet perform poorly on adversarially selected data. Prior work on adversarial inputs typically studies model oversensitivity: semantically invariant text perturbations that cause a model’s prediction to change. Here we focus on the complementary problem: excessive prediction undersensitivity, where input text is meaningfully changed but the model’s prediction does not, even though it should. We formulate an adversarial attack which searches among semantic variations of the question for which a model erroneously predicts the same answer, and with even higher probability. We demonstrate that models trained on both SQuAD2.0 and NewsQA are vulnerable to this attack, and then investigate data augmentation and adversarial training as defences. Both substantially decrease adversarial vulnerability, which generalises to held-out data and held-out attack spaces. Addressing undersensitivity furthermore improves model robustness on the previously introduced ADDSENT and ADDONESENT datasets, and models generalise better when facing train / evaluation distribution mismatch: they are less prone to overly rely on shallow predictive cues present only in the training set, and outperform a conventional model by as much as 10.9% F1.
Anthology ID:
2020.findings-emnlp.103
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2020
Month:
November
Year:
2020
Address:
Online
Venues:
EMNLP | Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1152–1165
Language:
URL:
https://aclanthology.org/2020.findings-emnlp.103
DOI:
10.18653/v1/2020.findings-emnlp.103
Bibkey:
Cite (ACL):
Johannes Welbl, Pasquale Minervini, Max Bartolo, Pontus Stenetorp, and Sebastian Riedel. 2020. Undersensitivity in Neural Reading Comprehension. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 1152–1165, Online. Association for Computational Linguistics.
Cite (Informal):
Undersensitivity in Neural Reading Comprehension (Welbl et al., Findings 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.findings-emnlp.103.pdf
Data
NewsQA