I’m an MSCS student at the University of Southern California.

Previously I was a research intern at BAIR, UC Berkeley, working on security for large language models. I was really fortunate to be advised by Prof. David Wagner and Postdoc Zhanhao Hu. Before that I spent time as a research intern in the CSE Department at HKUST, advised by Prof. Qian Zhang, working on backdoor defenses and human activity recognition. I obtained my B.Eng. degree at the School of Computer Science, Wuhan University.

I’m broadly interested in research topics related to LLMs, including alignment, RL, agentic RL and so on. More concretely, most of my work so far has asked where safety-aligned models break down — and what a cheap, practical defense looks like once you know.

You can find my full CV here: Jesson’s Curriculum Vitae.

News

• Jan. 2026 — JULI: Jailbreak Large Language Models by Self-Introspection accepted at ICLR.
• Mar. 2025 — MobHAR published in IMWUT (Vol. 9, Issue 1).
• Oct. 2024 — ARTEMIS accepted at IEEE Transactions on Dependable and Secure Computing.

Contact

The best way to reach me is by email at jessonwong1@gmail.com.