ML Security PapersShallowJail: Steering Jailbreaks against Large Language Models
Shang Liu, Hanyu Pei, Zeyan Liu · University of Louisville
Exploits shallow LLM alignment by injecting activation steering vectors into hidden states, achieving >90% jailbreak success without gradient optimization
Large Language Models(LLMs) have been successful in numerous fields. Alignment has usually been applied to prevent them from harmful purposes. However, aligned LLMs remain vulnerable to jailbreak attacks that deliberately mislead them into producing harmful outputs. Existing jailbreaks are either black-box, using carefully crafted, unstealthy prompts, or white-box, requiring resource-intensive computation. In light of these challenges, we introduce ShallowJail, a novel attack that exploits shallow alignment in LLMs. ShallowJail can misguide LLMs' responses by manipulating the initial tokens during inference. Through extensive experiments, we demonstrate the effectiveness of ShallowJail, which substantially degrades the safety of state-of-the-art LLM responses. Our code is available at https://github.com/liuup/ShallowJail.