COLING 2022main12 citations

Don’t Judge a Language Model by Its Last Layer: Contrastive Learning with Layer-Wise Attention Pooling

Dongsuk Oh, Yejin Kim, Hodong Lee, H. Howie Huang, Heuiseok Lim

Abstract

Recent pre-trained language models (PLMs) achieved great success on many natural language processing tasks through learning linguistic features and contextualized sentence representation. Since attributes captured in stacked layers of PLMs are not clearly identified, straightforward approaches such as embedding the last layer are commonly preferred to derive sentence representations from PLMs. This paper introduces the attention-based pooling strategy, which enables the model to preserve layer-wise signals captured in each layer and learn digested linguistic features for downstream tasks. The contrastive learning objective can adapt the layer-wise attention pooling to both unsupervised and supervised manners. It results in regularizing the anisotropic space of pre-trained embeddings and being more uniform. We evaluate our model on standard semantic textual similarity (STS) and semantic search tasks. As a result, our method improved the performance of the base contrastive learned BERTbase and variants.

BibTeX
@inproceedings{oh-etal-2022-dont,
    title = "Don{'}t Judge a Language Model by Its Last Layer: Contrastive Learning with Layer-Wise Attention Pooling",
    author = "Oh, Dongsuk  and
      Kim, Yejin  and
      Lee, Hodong  and
      Huang, H. Howie  and
      Lim, Heuiseok",
    editor = "Calzolari, Nicoletta  and
      Huang, Chu-Ren  and
      Kim, Hansaem  and
      Pustejovsky, James  and
      Wanner, Leo  and
      Choi, Key-Sun  and
      Ryu, Pum-Mo  and
      Chen, Hsin-Hsi  and
      Donatelli, Lucia  and
      Ji, Heng  and
      Kurohashi, Sadao  and
      Paggio, Patrizia  and
      Xue, Nianwen  and
      Kim, Seokhwan  and
      Hahm, Younggyun  and
      He, Zhong  and
      Lee, Tony Kyungil  and
      Santus, Enrico  and
      Bond, Francis  and
      Na, Seung-Hoon",
    booktitle = "Proceedings of the 29th International Conference on Computational Linguistics",
    month = oct,
    year = "2022",
    address = "Gyeongju, Republic of Korea",
    publisher = "International Committee on Computational Linguistics",
    url = "https://aclanthology.org/2022.coling-1.405/",
    pages = "4585--4592"
}
Don’t Judge a Language Model by Its Last Layer: Contrastive Learning with Layer-Wise Attention Pooling · COLING 2022