3 papers
cs.CL2025
Forewarned is Forearmed: Pre-Synthesizing Jailbreak-like Instructions to Enhance LLM Safety Guardrail to Potential Attacks
Sheng Liu, Qiang Sheng, Danding Wang +3
Despite advances in improving large language model (LLM) to refuse to answer malicious instructions, widely used LLMs remain vulnerable to jailbreak attacks where attackers generat…
cs.CL2025
The Staircase of Ethics: Probing LLM Value Priorities through Multi-Step Induction to Complex Moral Dilemmas
Ya Wu, Qiang Sheng, Danding Wang +5
Ethical decision-making is a critical aspect of human judgment, and the growing use of LLMs in decision-support systems necessitates a rigorous evaluation of their moral reasoning…
cs.CL2024
Exploiting User Comments for Early Detection of Fake News Prior to Users' Commenting
Qiong Nan, Qiang Sheng, Juan Cao +4
Both accuracy and timeliness are key factors in detecting fake news on social media. However, most existing methods encounter an accuracy-timeliness dilemma: Content-only methods g…