profile.jpg

johnjaejunlee95's Cozy AI Blog

Ph.D. Candidate @ Graduate School of Artificial Intelligence, UNIST

I am a fifth-year Ph.D. candidate whose research focuses on understanding how AI models learn and generalize from a theoretical perspective. My previous work centered on meta-learning, while my current research interests lie in multimodal learning, robustness, and generalization. More recently, I have also been exploring AI applications for biomedical data.

Meta-Learning Generalization & Robustness Multimodal Learning Biomedical AI

selected publications [full list]

(*) denotes equal contribution

  1. ICML
    Understanding Multimodal Learning: A Loss Landscape Smoothness Perspective
    Jae-Jun Lee, and Sung Whan Yoon
    In Forty-third International Conference on Machine Learning (ICML)
    Paper Code
  2. ICLR
    Can One Modality Model Synergize Training of Other Modality Models?
    Jae-Jun Lee, and Sung Whan Yoon
    In The Thirteenth International Conference on Learning Representations (ICLR)
    Paper Code
  3. AISTATS
    XB-MAML: Learning Expandable Basis Parameters for Effective Meta-Learning with Wide Task Coverage
    Jae-Jun Lee, and Sung Whan Yoon
    In International Conference on Artificial Intelligence and Statistics (AISTATS)
    Paper Code