Hi, I'm Hantao Lou.
I am a final-year undergraduate at Tong Class, Peking University, and a visiting researcher at UC Berkeley RDI, working on formal verification and specifications with Prof. Dawn Song.
My Hamming Question
How can we make advanced AI systems reliable and scalable?
I build supervision and verification methods that make complex model and agent behavior easier to inspect, evaluate, and trust.
Recent work
Selected highlights
FormalJudge: neuro-symbolic agentic oversight
Language Models Resist Alignment
Experience
Where I have worked
UC Berkeley RDI · Visiting Researcher
Formal verification and specifications with Prof. Dawn Song.
Peking University · Undergraduate Researcher
Scalable AI alignment with Prof. Yaodong Yang.
MATS · Research Scholar
Mechanistic interpretability with Evan Hubinger.
Beneficial AI Foundation
Formal verification supervised by Prof. Max Tegmark.