Year: 2026
Type: conference-paper
Abstract: Large Language Models (LLMs) have achieved substantial progress in alignment, ensuring safer and more reliable outputs.However, jailbreak attacks can still bypass these safeguards and provoke harmful responses from well-aligned models.While some studies have achieved defenses against jailbreak attacks by modifying output distributions or detecting harmful content, the exact rationale still remains elusive.In this work, we present a novel neuron-level interpretability method that focuses on the role of safety-related knowledge neurons.Unlike existing approaches, our method projects the model's internal representation into a more consistent and interpretable vocabulary space.We then show that adjusting the activation of safety-related neurons can effectively control the models behavior with a mean ASR higher than 97%.Building on this insight, we propose SafeTuning, a fine-tuning strategy that reinforces safety-critical neurons to improve model robustness against jailbreaks.Safe-Tuning consistently reduces attack success rates across multiple LLMs and outperforms all four baseline defenses.These findings offer a new perspective on understanding and defending against jailbreak attacks.Warning: this paper may contain offensive prompts and outputs.
Authors: Chongwen Zhao, Yutong Ke, Kaizhu Huang
Institution: Duke Kunshan University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Awards: G1698772872, G5866557873
OpenAlex ID: W7140123232