Discourse-Aware Soft Prompting for Text Generation

Marjan Ghazvininejad, Vladimir Karpukhin, Vera Gor, Asli Celikyilmaz


Abstract
Current efficient fine-tuning methods(e.g., adapters, prefix-tuning, etc.) have optimized conditional text generation via training a small set of extra parameters of the neural language model, while freezing the rest for efficiency. While showing strong performance on some generation tasks, they don’t generalize across all generation tasks. We show that soft-prompt based conditional text generation can be improved with simple and efficient methods that simulate modeling the discourse structure of human written text.We investigate two design choices: First, we apply hierarchical blocking on the prefix parameters to simulate a higher-level discourse structure of human written text. Second, we apply attention sparsity on the prefix parameters at different layers of the network and learn sparse transformations on the softmax-function. We show that structured design of prefix parameters yields more coherent, faithful and relevant generations than the baseline prefix-tuning on all generation tasks.
Anthology ID:
2022.emnlp-main.303
Volume:
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
4570–4589
Language:
URL:
https://aclanthology.org/2022.emnlp-main.303
DOI:
10.18653/v1/2022.emnlp-main.303
Bibkey:
Cite (ACL):
Marjan Ghazvininejad, Vladimir Karpukhin, Vera Gor, and Asli Celikyilmaz. 2022. Discourse-Aware Soft Prompting for Text Generation. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 4570–4589, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Discourse-Aware Soft Prompting for Text Generation (Ghazvininejad et al., EMNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.emnlp-main.303.pdf