About
I am a Ph.D. student in Artificial Intelligence at HKUST (Guangzhou), advised by Prof. Lionel Ni. Previously, I was a MPhil student in Data Science at HKUST, advised by Prof. Qiong Luo and Prof. Yang Xiang. My research interests lie at the intersection of Agentic Systems, Trustworthy AI, and Computational Neuroscience. I focus on scaling and orchestrating large multi-agent systems, securing agentic AI against jailbreaks, sandbox escapes, and tool-permission bypasses, and developing multi-agent systems for neuroscience research.
News
Full CV & timeline →-
Started a research internship at Harvard Medical School, exploring hypothesis ranking with human–LLM co-evaluation and wet-lab validation.
-
Joined the University of Hong Kong NLP Lab as a research assistant, working on automatic agent swarms for future prediction.
- Two paper accepted at ACL 2026 Findings [ConflictHarm, Select2Reason] .
-
One paper accepted at EACL 2026 Findings [FlowHarm].
Selected work
All publications →-
Conflicts Make Large Reasoning Models Vulnerable to Attacks
Findings of ACL 2026 · pp. 9509–9531 · San Diego, California, USA
-
Continual Pretraining on Encrypted Synthetic Data for Privacy-Preserving LLMs
Findings of EACL 2026 · pp. 408–425 · Rabat, Morocco
-
