Making Pre-trained Language Models both Task-solvers and Self-calibrators

Yangyi Chen, Xingyao Wang, Heng Ji


Abstract
Pre-trained language models (PLMs) serve as backbones for various real-world systems. For high-stake applications, it’s equally essential to have reasonable confidence estimations in predictions. While the vanilla confidence scores of PLMs can already be effectively utilized, PLMs consistently become overconfident in their wrong predictions, which is not desirable in practice. Previous work shows that introducing an extra calibration task can mitigate this issue. The basic idea involves acquiring additional data to train models in predicting the confidence of their initial predictions. However, it only demonstrates the feasibility of this kind of method, assuming that there are abundant extra available samples for the introduced calibration task. In this work, we consider the practical scenario that we need to effectively utilize training samples to make PLMs both task-solvers and self-calibrators. Three challenges are presented, including limited training samples, data imbalance, and distribution shifts. We first conduct pilot experiments to quantify various decisive factors in the calibration task. Based on the empirical analysis results, we propose a training algorithm LM-TOAST to tackle the challenges. Experimental results show that LM-TOAST can effectively utilize the training data to make PLMs have reasonable confidence estimations while maintaining the original task performance. Further, we consider three downstream applications, namely selective classification, adversarial defense, and model cascading, to show the practical usefulness of LM-TOAST.
Anthology ID:
2023.findings-acl.624
Volume:
Findings of the Association for Computational Linguistics: ACL 2023
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
9845–9862
Language:
URL:
https://aclanthology.org/2023.findings-acl.624
DOI:
10.18653/v1/2023.findings-acl.624
Bibkey:
Cite (ACL):
Yangyi Chen, Xingyao Wang, and Heng Ji. 2023. Making Pre-trained Language Models both Task-solvers and Self-calibrators. In Findings of the Association for Computational Linguistics: ACL 2023, pages 9845–9862, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
Making Pre-trained Language Models both Task-solvers and Self-calibrators (Chen et al., Findings 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/add_acl24_videos/2023.findings-acl.624.pdf