LexFit: Lexical Fine-Tuning of Pretrained Language Models

Ivan Vulić, Edoardo Maria Ponti, Anna Korhonen, Goran Glavaš


Abstract
Transformer-based language models (LMs) pretrained on large text collections implicitly store a wealth of lexical semantic knowledge, but it is non-trivial to extract that knowledge effectively from their parameters. Inspired by prior work on semantic specialization of static word embedding (WE) models, we show that it is possible to expose and enrich lexical knowledge from the LMs, that is, to specialize them to serve as effective and universal “decontextualized” word encoders even when fed input words “in isolation” (i.e., without any context). Their transformation into such word encoders is achieved through a simple and efficient lexical fine-tuning procedure (termed LexFit) based on dual-encoder network structures. Further, we show that LexFit can yield effective word encoders even with limited lexical supervision and, via cross-lingual transfer, in different languages without any readily available external knowledge. Our evaluation over four established, structurally different lexical-level tasks in 8 languages indicates the superiority of LexFit-based WEs over standard static WEs (e.g., fastText) and WEs from vanilla LMs. Other extensive experiments and ablation studies further profile the LexFit framework, and indicate best practices and performance variations across LexFit variants, languages, and lexical tasks, also directly questioning the usefulness of traditional WE models in the era of large neural models.
Anthology ID:
2021.acl-long.410
Volume:
Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)
Month:
August
Year:
2021
Address:
Online
Editors:
Chengqing Zong, Fei Xia, Wenjie Li, Roberto Navigli
Venues:
ACL | IJCNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5269–5283
Language:
URL:
https://aclanthology.org/2021.acl-long.410
DOI:
10.18653/v1/2021.acl-long.410
Bibkey:
Cite (ACL):
Ivan Vulić, Edoardo Maria Ponti, Anna Korhonen, and Goran Glavaš. 2021. LexFit: Lexical Fine-Tuning of Pretrained Language Models. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 5269–5283, Online. Association for Computational Linguistics.
Cite (Informal):
LexFit: Lexical Fine-Tuning of Pretrained Language Models (Vulić et al., ACL-IJCNLP 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.acl-long.410.pdf
Video:
 https://aclanthology.org/2021.acl-long.410.mp4