Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA

Yichen Jiang, Mohit Bansal


Abstract
Multi-hop question answering requires a model to connect multiple pieces of evidence scattered in a long context to answer the question. In this paper, we show that in the multi-hop HotpotQA (Yang et al., 2018) dataset, the examples often contain reasoning shortcuts through which models can directly locate the answer by word-matching the question with a sentence in the context. We demonstrate this issue by constructing adversarial documents that create contradicting answers to the shortcut but do not affect the validity of the original answer. The performance of strong baseline models drops significantly on our adversarial test, indicating that they are indeed exploiting the shortcuts rather than performing multi-hop reasoning. After adversarial training, the baseline’s performance improves but is still limited on the adversarial test. Hence, we use a control unit that dynamically attends to the question at different reasoning hops to guide the model’s multi-hop reasoning. We show that our 2-hop model trained on the regular data is more robust to the adversaries than the baseline. After adversarial training, it not only achieves significant improvements over its counterpart trained on regular data, but also outperforms the adversarially-trained baseline significantly. Finally, we sanity-check that these improvements are not obtained by exploiting potential new shortcuts in the adversarial data, but indeed due to robust multi-hop reasoning skills of the models.
Anthology ID:
P19-1262
Volume:
Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2019
Address:
Florence, Italy
Editors:
Anna Korhonen, David Traum, Lluís Màrquez
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2726–2736
Language:
URL:
https://aclanthology.org/P19-1262
DOI:
10.18653/v1/P19-1262
Bibkey:
Cite (ACL):
Yichen Jiang and Mohit Bansal. 2019. Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 2726–2736, Florence, Italy. Association for Computational Linguistics.
Cite (Informal):
Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA (Jiang & Bansal, ACL 2019)
Copy Citation:
PDF:
https://aclanthology.org/P19-1262.pdf
Video:
 https://aclanthology.org/P19-1262.mp4
Code
 jiangycTarheel/Adversarial-MultiHopQA
Data
HotpotQASQuAD