Jeongyeon Hwang

Jeongyeon Hwang

Jeongyeon Hwang

Hi! I’m Jeongyeon Hwang, a fourth year integrated M.S./Ph.D. candidate in the Graduate School of Artificial Intelligence at POSTECH, advised by Jungseul Ok.

I’m interested in what goes wrong when AI systems meet the real world. Existing evaluations often assume clean, idealized conditions—but in deployment, things break down across many surfaces: inputs, outputs, and weights. I study these failures under realistic threat models and build safeguards that hold up in actual deployment, working toward a zero-trust approach to AI we can rely on.

Please feel free to reach out via email at {firstname}.{lastname}@postech.ac.kr.

News

  • Jun 2026Selected as a Doctoral Excellence Scholarship recipient by KOSAF.
  • May 2026A paper (BIRA) got accepted at ICML 2026.
  • Jan 2026Received an Honorable Mention at the BK21 Outstanding Paper Awards from POSTECH GSAI.
  • Sep 2025Beginning a research visit at NYU (September–December).
  • Aug 2025Two papers (RA-RAG, LSC) got accepted at EMNLP 2025.
  • Aug 2024A paper (MedBN) got accepted at CVPR 2024.

Selected Publications

LLM Watermark Evasion via Bias Inversion

Jeongyeon Hwang, Sangdon Park, Jungseul Ok

ICML 2026

Paper

Efficient Latent Semantic Clustering for Scaling Test-Time Computation of LLMs

Sungjae Lee, Hoyoung Kim, Jeongyeon Hwang, Eunhyeok Park, Jungseul Ok

EMNLP, 2025 Findings (long)

Paper

Retrieval-Augmented Generation with Estimation of Source Reliability

Jeongyeon Hwang, Junyoung Park, Hyejin Park, Dongwoo Kim, Sangdon Park, Jungseul Ok

EMNLP, 2025 Main (long)

Paper

MedBN: Robust Test-Time Adaptation against Malicious Test Samples

Hyejin Park*, Jeongyeon Hwang*, Sunung Mun, Sangdon Park, Jungseul Ok

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024.

Paper