Real-World Compositional Generalization with Disentangled Sequence-to-Sequence Learning

Hao Zheng, Mirella Lapata


Abstract
Compositional generalization is a basic mechanism in human language learning, which current neural networks struggle with. A recently proposed Disentangled sequence-to-sequence model (Dangle) shows promising generalization capability by learning specialized encodings for each decoding step. We introduce two key modifications to this model which encourage more disentangled representations and improve its compute and memory efficiency, allowing us to tackle compositional generalization in a more realistic setting. Specifically, instead of adaptively re-encoding source keys and values at each time step, we disentangle their representations and only re-encode keys periodically, at some interval. Our new architecture leads to better generalization performance across existing tasks and datasets, and a new machine translation benchmark which we create by detecting naturally occurring compositional patterns in relation to a training set. We show this methodology better emulates real-world requirements than artificial challenges.
Anthology ID:
2023.findings-acl.108
Volume:
Findings of the Association for Computational Linguistics: ACL 2023
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1711–1725
Language:
URL:
https://aclanthology.org/2023.findings-acl.108
DOI:
10.18653/v1/2023.findings-acl.108
Bibkey:
Cite (ACL):
Hao Zheng and Mirella Lapata. 2023. Real-World Compositional Generalization with Disentangled Sequence-to-Sequence Learning. In Findings of the Association for Computational Linguistics: ACL 2023, pages 1711–1725, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
Real-World Compositional Generalization with Disentangled Sequence-to-Sequence Learning (Zheng & Lapata, Findings 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.findings-acl.108.pdf