Sowing the Wind, Reaping the Whirlwind: The Impact of Editing Language Models

Rima Hazra, Sayan Layek, Somnath Banerjee, Soujanya Poria

Abstract
In the rapidly advancing field of artificial intelligence, the concept of ‘Red-Teaming’ or ‘Jailbreaking’ large language models (LLMs) has emerged as a crucial area of study. This approach is especially significant in terms of assessing and enhancing the safety and robustness of these models. This paper investigates the intricate consequences of such modifications through model editing, uncovering a complex relationship between enhancing model accuracy and preserving its ethical integrity. Our in-depth analysis reveals a striking paradox: while injecting accurate information is crucial for model reliability, it can paradoxically destabilize the model’s foundational framework, resulting in unpredictable and potentially unsafe behaviors. Additionally, we propose a benchmark dataset NicheHazardQA to investigate this unsafe behavior both within the same and cross topical domain. This aspect of our research sheds light on how the edits, impact the model’s safety metrics and guardrails. Our findings show that model editing serves as a cost-effective tool for topical red-teaming by methodically applying targeted edits and evaluating the resultant model behavior.
Anthology ID:
2024.findings-acl.960
Volume:
Findings of the Association for Computational Linguistics ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand and virtual meeting
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
16227–16239
Language:
URL:
https://aclanthology.org/2024.findings-acl.960
DOI:
Bibkey:
Cite (ACL):
Rima Hazra, Sayan Layek, Somnath Banerjee, and Soujanya Poria. 2024. Sowing the Wind, Reaping the Whirlwind: The Impact of Editing Language Models. In Findings of the Association for Computational Linguistics ACL 2024, pages 16227–16239, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics.
Cite (Informal):
Sowing the Wind, Reaping the Whirlwind: The Impact of Editing Language Models (Hazra et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2024.findings-acl.960.pdf
PDF Search