The CRINGE Loss: Learning what language not to model

Leonard Adolphs, Tianyu Gao, Jing Xu, Kurt Shuster, Sainbayar Sukhbaatar, Jason Weston


Abstract
Standard language model training employs gold human documents or human-human interaction data, and treats all training data as positive examples. Growing evidence shows that even with very large amounts of positive training data, issues remain that can be alleviated with relatively small amounts of negative data – examples of what the model should not do. In this work, we propose a novel procedure to train with such data called the “CRINGE” loss (ContRastive Iterative Negative GEneration). We show the effectiveness of this approach across three different experiments on the tasks of safe generation, contradiction avoidance, and open-domain dialogue. Our models outperform multiple strong baselines and are conceptually simple, easy to train and implement.
Anthology ID:
2023.acl-long.493
Volume:
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
8854–8874
Language:
URL:
https://aclanthology.org/2023.acl-long.493
DOI:
10.18653/v1/2023.acl-long.493
Bibkey:
Cite (ACL):
Leonard Adolphs, Tianyu Gao, Jing Xu, Kurt Shuster, Sainbayar Sukhbaatar, and Jason Weston. 2023. The CRINGE Loss: Learning what language not to model. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 8854–8874, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
The CRINGE Loss: Learning what language not to model (Adolphs et al., ACL 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/landing_page/2023.acl-long.493.pdf
Video:
 https://preview.aclanthology.org/landing_page/2023.acl-long.493.mp4