No Offense Taken: Eliciting Offensiveness from Language Models
This work addresses the need for safe deployment of language models by providing automated tools to test for offensiveness, though it is incremental as it builds on existing red teaming methods.
The paper tackles the problem of robustly testing language models for offensiveness by developing an automated pipeline for generating diverse test cases, resulting in a corpus that helps elicit offensive responses and identify failure modes in widely deployed models.
This work was completed in May 2022. For safe and reliable deployment of language models in the real world, testing needs to be robust. This robustness can be characterized by the difficulty and diversity of the test cases we evaluate these models on. Limitations in human-in-the-loop test case generation has prompted an advent of automated test case generation approaches. In particular, we focus on Red Teaming Language Models with Language Models by Perez et al.(2022). Our contributions include developing a pipeline for automated test case generation via red teaming that leverages publicly available smaller language models (LMs), experimenting with different target LMs and red classifiers, and generating a corpus of test cases that can help in eliciting offensive responses from widely deployed LMs and identifying their failure modes.