HistAlign: Improving Context Dependency in Language Generation by Aligning with History

David Wan, Shiyue Zhang, Mohit Bansal


Abstract
Language models (LMs) can generate hallucinations and incoherent outputs, which highlights their weak context dependency. Cache-LMs, which augment LMs with a memory of recent history, can increase context dependency and have shown remarkable performance in diverse language generation tasks. However, we find that even with training, the performance gain stemming from the cache component of current cache-LMs is suboptimal due to the misalignment between the current hidden states and those stored in the memory. In this work, we present HistAlign, a new training approach to ensure good cache alignment such that the model receives useful signals from the history. We first prove our concept on a simple and synthetic task where the memory is essential for correct predictions, and we show that the cache component of HistAlign is better aligned and improves overall performance. Next, we evaluate HistAlign on diverse downstream language generation tasks, including prompt continuation, abstractive summarization, and data-to-text. We demonstrate that HistAlign improves text coherence and faithfulness in open-ended and conditional generation settings respectively. HistAlign is also generalizable across different model families, showcasing its strength in improving context dependency of LMs in diverse scenarios.
Anthology ID:
2023.emnlp-main.179
Volume:
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2023
Address:
Singapore
Editors:
Houda Bouamor, Juan Pino, Kalika Bali
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2941–2960
Language:
URL:
https://aclanthology.org/2023.emnlp-main.179
DOI:
10.18653/v1/2023.emnlp-main.179
Bibkey:
Cite (ACL):
David Wan, Shiyue Zhang, and Mohit Bansal. 2023. HistAlign: Improving Context Dependency in Language Generation by Aligning with History. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 2941–2960, Singapore. Association for Computational Linguistics.
Cite (Informal):
HistAlign: Improving Context Dependency in Language Generation by Aligning with History (Wan et al., EMNLP 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.emnlp-main.179.pdf
Video:
 https://aclanthology.org/2023.emnlp-main.179.mp4