Jiachen Ma (马嘉晨)

I am a first-year Ph.D. student at Fudan University and Shanghai Artificial Intelligence Laboratory (PJLab), where I am privileged to be advised by Prof. Bowen Zhou and Prof. Chao Yang.

Prior to my doctoral studies, I earned my M.S. from Zhejiang University in 2025.

My research is driven by a fundamental commitment to Trustworthy Machine Learning. I am dedicated to exploring the boundaries of AI safety and ensuring that the next generation of intelligence remains beneficial and controllable.

🔍 Research Interests

I am broadly interested in the intersection of safety and general intelligence. Currently, I am dedicated to the following directions:

• Value Alignment: Designing mechanisms that ensure large-scale models internalize human values and safety constraints fundamentally.
• Autonomous Agents: Investigating the safety alignment and red teaming of agents in complex, long-horizon decision-making environments.
• Explainability of Model Safety: Uncovering the internal mechanisms and representation-level dynamics that govern model (mis)behaviors to build more transparent safety frameworks.

“My vision is to make Safe AI — ensuring safety is not just a constraint, but a foundational property of intelligence.”