I am an M.S. student in the Mobile Intelligence & Interaction Lab at KAIST, advised by Professor Sung-Ju Lee. I received my B.S. in Electrical Engineering from KAIST, graduating summa cum laude.

My research interests include reliable clinical foundation models and the faithfulness of model explanations. I currently lead a project investigating whether explanations from ECG language models help predict their behavior on related ECGs, beyond the information provided by the original answer.

CV Email LinkedIn GitHub

News

Attended ICML 2026 in Seoul :kr:.
Presented a poster at NeurIPS 2025 in San Diego :us:.
Selected as a recipient of the Kwon Young-se Scholarship (KAIST EE).
Paper accepted at NeurIPS 2025.
Attended UIST 2025 in Busan :kr:.
Started my M.S. at KAIST EE.
Graduated from KAIST with a B.S. in Electrical Engineering.
Visiting student at Rice University.

Publications

  1. MedMix: Specialization-Consistent Federated Sparse MoEs under Modality Heterogeneity
    Adiba Orzikulova, Dong Min Kim, Jaehong Yoon, and Sung-Ju Lee
    arXiv preprint, 2026
    Paper
  2. SNAP: Low-Latency Test-Time Adaptation with Sparse Updates
    Hyeongheon Cha, Dong Min Kim, Hye Won Chung, Taesik Gong, and Sung-Ju Lee
    NeurIPS 2025
    Website Paper Code