Dodrio: Exploring Transformer Models with Interactive Visualization

Zijie J. Wang, Robert Turko, Duen Horng Chau


Abstract
Why do large pre-trained transformer-based models perform so well across a wide variety of NLP tasks? Recent research suggests the key may lie in multi-headed attention mechanism’s ability to learn and represent linguistic information. Understanding how these models represent both syntactic and semantic knowledge is vital to investigate why they succeed and fail, what they have learned, and how they can improve. We present Dodrio, an open-source interactive visualization tool to help NLP researchers and practitioners analyze attention mechanisms in transformer-based models with linguistic knowledge. Dodrio tightly integrates an overview that summarizes the roles of different attention heads, and detailed views that help users compare attention weights with the syntactic structure and semantic information in the input text. To facilitate the visual comparison of attention weights and linguistic knowledge, Dodrio applies different graph visualization techniques to represent attention weights scalable to longer input text. Case studies highlight how Dodrio provides insights into understanding the attention mechanism in transformer-based models. Dodrio is available at https://poloclub.github.io/dodrio/.
Anthology ID:
2021.acl-demo.16
Volume:
Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing: System Demonstrations
Month:
August
Year:
2021
Address:
Online
Editors:
Heng Ji, Jong C. Park, Rui Xia
Venues:
ACL | IJCNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
132–141
Language:
URL:
https://aclanthology.org/2021.acl-demo.16
DOI:
10.18653/v1/2021.acl-demo.16
Bibkey:
Cite (ACL):
Zijie J. Wang, Robert Turko, and Duen Horng Chau. 2021. Dodrio: Exploring Transformer Models with Interactive Visualization. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing: System Demonstrations, pages 132–141, Online. Association for Computational Linguistics.
Cite (Informal):
Dodrio: Exploring Transformer Models with Interactive Visualization (Wang et al., ACL-IJCNLP 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.acl-demo.16.pdf
Optional supplementary material:
 2021.acl-demo.16.OptionalSupplementaryMaterial.zip
Video:
 https://aclanthology.org/2021.acl-demo.16.mp4
Code
 poloclub/dodrio
Data
GLUEPenn TreebankSSTSST-2