Juho Lee

Juho Lee is a faculty member in the Department of Computer Science at KAIST. His research is associated with the Kim Jaechul Graduate School of AI. He is affiliated with KAIST and can be contacted via email at juholee@kaist.ac.kr. His academic profile includes links to Google Scholar, GitHub, and a CV. His work is organized into publications grouped by type, with each list sorted by date. He also engages in teaching, talks, and service activities.

Juho Lee is an associate professor at KAIST, specializing in Bayesian deep learning and Bayesian nonparametric models. His research focuses on developing probabilistic frameworks for neural networks to handle uncertainty and improve interpretability. Key areas include attention mechanisms for permutation-invariant models, generative modeling of structured data, and robustness against adversarial attacks. His work also explores uncertainty-aware methods for reliable prediction and efficient training strategies for large-scale models. Lee's publications address challenges in few-shot learning, time-series analysis, and safe self-distillation of diffusion models, emphasizing the integration of Bayesian principles with deep learning techniques.