Bo-Han Lai

賴柏翰

🙋🏻‍♂️ ML Researcher
Taipei, Taiwan

About Me

I am currently a Ph.D. student in the Department of Computer Science and Information Engineering at National Taiwan University, conducting research with Prof. Shang-Tse Chen in the AISLab.

I'm interested in understanding how machine learning models acquire, represent, and utilize knowledge about the world, with the ultimate goal of building trustworthy, high-performing ML/AI systems to benefit humanity.

Please feel free to get in touch if you’re interested in collaboration or discussion 🤝

Selected Publications

Enhancing Certified Robustness via Block Reflector Orthgonal Layers & Logit Anealing Loss
Bo-Han Lai, 
Pin-Han Huang, 
Bo-Han Kung, 
Shang-Tse Chen
ICML 2025 (Spotlight)
May 2nd 2025
#Adversarial Robustness
#Lipschitz networks

We propose a low-rank, efficient orthogonal layer and a margin-balancing loss function to construct effective and certifiably robust neural networks.

ArXiv
GitHub
Project Page
Enhancing Certified Robustness via Block Reflector Orthgonal Layers & Logit Anealing Loss

Research Interests

Robustness in Machine Learning
Privacy-preserving Deep Learning
Machine Unlearning
Modular Model Adaptation
Last updated Oct 11th 2026