Learning diverse attacks on large language models for robust red-teaming and safety tuning
Reviewed by Pithpith:EGLX34FCopen to challenge →
read the original abstract
Red-teaming, or identifying prompts that elicit harmful responses, is a critical step in ensuring the safe and responsible deployment of large language models (LLMs). Developing effective protection against many modes of attack prompts requires discovering diverse attacks. Automated red-teaming typically uses reinforcement learning to fine-tune an attacker language model to generate prompts that elicit undesirable responses from a target LLM, as measured, for example, by an auxiliary toxicity classifier. We show that even with explicit regularization to favor novelty and diversity, existing approaches suffer from mode collapse or fail to generate effective attacks. As a flexible and probabilistically principled alternative, we propose to use GFlowNet fine-tuning, followed by a secondary smoothing phase, to train the attacker model to generate diverse and effective attack prompts. We find that the attacks generated by our method are effective against a wide range of target LLMs, both with and without safety tuning, and transfer well between target LLMs. Finally, we demonstrate that models safety-tuned using a dataset of red-teaming prompts generated by our method are robust to attacks from other RL-based red-teaming approaches.
This paper has not been read by Pith yet.
Forward citations
Cited by 2 Pith papers
-
Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balance
Stable-GFlowNet stabilizes GFN training for LLM red-teaming by eliminating Z estimation via pairwise comparisons and robust masking against noisy rewards while adding a fluency stabilizer.
-
Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balance
Stable-GFlowNet improves training stability and attack diversity in LLM red-teaming by eliminating Z estimation via contrastive trajectory balance while preserving GFN optimality.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.