Pingchuan Ma

机构：Hong Kong University of Science and Technology

发表论文 69 篇 · 总被引 1200 次 · h-index 21

代表论文

• Taming Various Privilege Escalation in LLM-Based Agent Systems: A Mandatory Access Control Framework (2026 · arXiv.org · 被引 50)
• Testing and Understanding Erroneous Planning in LLM Agents through Synthesized User Inputs (2024 · arXiv.org · 被引 29)
• SkillReducer: Optimizing LLM Agent Skills for Token Efficiency (2026 · arXiv.org · 被引 28)
• Measuring and Augmenting Large Language Models for Solving Capture-the-Flag Challenges (2025 · Conference on Computer and Communications Security · 被引 24)
• An Empirical Study on Large Language Models in Accuracy and Robustness under Chinese Industrial Scenarios (2024 · arXiv.org · 被引 11)
• Taxonomy, Evaluation and Exploitation of IPI-Centric LLM Agent Defense Frameworks (2025 · arXiv.org · 被引 7)