On the evolution of syntactic information encoded by BERT’s contextualized representations

Pérez-Mayos, Laura; Carlini, Roberto; Ballesteros, Miguel; Wanner, Leo

doi:http://dx.doi.org/10.18653/v1/2021.eacl-main.191

On the evolution of syntactic information encoded by BERT’s contextualized representations

Llicència

Citació

Pérez-Mayos L, Carlini R, Ballesteros M, Wanner L. On the evolution of syntactic information encoded by BERT’s contextualized representations. In: Merlo P, Tiedemann J, Tsarfaty R. The 16th Conference of the European Chapter of the Association for Computational Linguistics: proceedings of the Conference; 2021 Apr 19-23; online. Stroudsburg: Association for Computational Linguistics; 2021. p. 2243-58. DOI: 10.18653/v1/2021.eacl-main.191

Resum

The adaptation of pretrained language models to solve supervised tasks has become a baseline in NLP, and many recent works have focused on studying how linguistic information is encoded in the pretrained sentence representations. Among other information, it has been shown that entire syntax trees are implicitly embedded in the geometry of such models. As these models are often fine-tuned, it becomes increasingly important to understand how the encoded knowledge evolves along the fine-tuning. In this paper, we analyze the evolution of the embedded syntax trees along the fine-tuning process of BERT for six different tasks, covering all levels of the linguistic structure. Experimental results show that the encoded syntactic information is forgotten (PoS tagging), reinforced (dependency and constituency parsing) or preserved (semantics-related tasks) in different ways along the finetuning process depending on the task.

Descripció

Comunicació presentada a 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2021), celebrat del 19 al 23 d'abril de 2021 de manera virtual.