Post-Hoc Interpretation of Transformer Hyperparameters with Explainable Boosting Machines

Kiron Deb, Xuan Zhang, Kevin Duh


Abstract
Hyperparameter tuning is important for achieving high accuracy in deep learning models, yet little interpretability work has focused on hyperparameters. We propose to use the Explainable Boosting Machine (EBM), a glassbox method, as a post-hoc analysis tool for understanding how hyperparameters influence model accuracy. We present a case study on Transformer models in machine translation to illustrate the kinds of insights that may be gleaned, and perform extensive analysis to test the robustness of EBM under different data conditions.
Anthology ID:
2022.blackboxnlp-1.5
Volume:
Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates (Hybrid)
Editors:
Jasmijn Bastings, Yonatan Belinkov, Yanai Elazar, Dieuwke Hupkes, Naomi Saphra, Sarah Wiegreffe
Venue:
BlackboxNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
51–61
Language:
URL:
https://aclanthology.org/2022.blackboxnlp-1.5
DOI:
10.18653/v1/2022.blackboxnlp-1.5
Bibkey:
Cite (ACL):
Kiron Deb, Xuan Zhang, and Kevin Duh. 2022. Post-Hoc Interpretation of Transformer Hyperparameters with Explainable Boosting Machines. In Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, pages 51–61, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics.
Cite (Informal):
Post-Hoc Interpretation of Transformer Hyperparameters with Explainable Boosting Machines (Deb et al., BlackboxNLP 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.blackboxnlp-1.5.pdf