Hyesu Lim
Scholar

Hyesu Lim

Google Scholar ID: CqRG8v0AAAAJ
Korea Advanced Institute of Science and Technology (KAIST)
mechanistic interpretabilityrobustnesstrustworthiness
Citations & Impact
All-time
Citations
224
 
H-index
4
 
i10-index
4
 
Publications
12
 
Co-authors
5
list available
Resume (English only)
Academic Achievements
  • Paper accepted at NeurIPS 2025 (Sep 2025)
  • Paper accepted at MICCAI 2025 (May 2025)
  • Paper accepted at ICLR 2025 (Jan 2025)
  • Paper accepted at NeurIPS 2024 (Sep 2024)
  • Paper accepted at NeurIPS 2023 Workshop (Distshift) (Dec 2023)
  • Paper accepted at ICLR 2023 (Jan 2023)
  • Paper accepted at EMNLP 2021 (Aug 2021)
  • Preprint: 'ConceptScope: Characterizing Dataset Bias via Disentangled Visual Concepts' (Oct 2025)
  • Preprint: 'CytoSAE: Interpretable Cell Embeddings for Hematology' (Jul 2025)
  • Preprint: 'Sparse autoencoders reveal selective remapping of visual concepts during adaptation' (Dec 2024)
  • Selected as student volunteer for MICCAI 2025 (Jul 2025)
Research Experience
  • Research collaboration at Helmholtz Munich with Dr. Steffen Schneider
  • Research intern at NAVER AI Lab with Dr. Dongyoon Han (since Oct 2024)
  • Research intern at Qualcomm AI Research with Dr. Sungha Choi (since Feb 2022)
  • Visiting scholar at Carnegie Mellon University (since Aug 2023), supported by IITP scholarship
  • Summer Institute Programme at Helmholtz Munich (Jul 2024), supported by NRF/DAAD scholarship