Ha Manh Bui

Hi! I am a Ph.D. Candidate in Computer Science at Johns Hopkins University, advised by Angie Liu, and also fortunate to collaborate with Eric Nalisnick, Enrique Mallada, and Kimia Ghodabi. Prior to that, I was at VinAI (Qualcomm AI) and obtained a B.S.E. in Information Technology from PTIT-Hanoi, under the supervision of Toan Tran and Dinh Phung.

Research Interests: I am interested in reliable Artificial Intelligence (AI) models that can support people in high-stakes applications (e.g., healthcare, finance, robotics). To accomplish this, my research focuses on developing principled Machine Learning methods for:

Uncertainty & Robustness: enhancing the efficiency (speed, memory), uncertainty (model's confidence), and robustness (generalization under distribution shifts) of modern AI models.
Reinforcement Learning: utilizing robust & uncertainty-aware AI models for learning to make good decisions (exploration, exploitation) in sequential problems.

Publications

Efficient and Uncertainty-Aware Diffusion Framework for Offline-to-Online Reinforcement Learning [] Ha Manh Bui, Metod Jazbec, Eric Nalisnick, Anqi Liu International Conference on Machine Learning, 2026

Q-Learning with Shift-Aware Upper Confidence Bound in Non-Stationary Reinforcement Learning [PDF, article, code] Ha Manh Bui, Felix Parker, Kimia Ghobadi, Anqi Liu International Conference on Artificial Intelligence and Statistics, 2026

Variance-Aware Linear UCB with Deep Representation for Neural Contextual Bandits [PDF, article, Colab demo, code] Ha Manh Bui, Enrique Mallada, Anqi Liu International Conference on Artificial Intelligence and Statistics, 2025

Density-Softmax: Efficient Test-time Model for Uncertainty Estimation and Robustness under Distribution Shifts [PDF, article, Colab demo, code] Ha Manh Bui, Anqi Liu International Conference on Machine Learning, 2024

Density-Regression: Efficient and Distance-Aware Deep Regressor for Uncertainty Estimation under Distribution Shifts [PDF, article, Colab demo, code] Ha Manh Bui, Anqi Liu International Conference on Artificial Intelligence and Statistics, 2024

Exploiting Domain-Specific Features to Enhance Domain Generalization [PDF, article, code] Manh-Ha Bui, Toan Tran, Anh Tran, Dinh Phung Advances in Neural Information Processing Systems, 2021