cv
This is a description of the page. You can modify it in '_pages/cv.md'. You can also change or remove the top pdf download button.
Basics
| Name | Yong-Hyun Park |
| Label | Incoming PhD Student |
| enkeejunior1@snu.ac.kr | |
| Phone | +82 10 8841 9051 |
| Summary | Incoming PhD student at University of Pennsylvania (Fall 2025). Experienced in generative models, diffusion models, and safe text-to-image generation. |
Work
-
2024.09 - Present Research Intern (Music Foundation Model, Deep Generative Model)
SONY AI
Improving the sampling algorithm of discrete diffusion models.
-
2024.03 - 2024.08 Research Intern (Generation Research)
NAVER AI Lab
Enhancing the safety of text-to-image models through unlearning algorithms.
-
2020.10 - 2021.08 Undergraduate Research Intern
CCS Lab, Seoul National University
Developed Bayesian model (Hierarchical Gaussian Filter) for addiction behavior.
Education
-
2025.09 - Present Philadelphia, USA
-
2022.09 - 2024.08 Seoul, South Korea
MS
Seoul National University
Science Education (Physics Major)
- Mechanics and Education
- Electromagnetism and Education
- Quantum Physics and Education
- Statistical Physics and Education
- Mathematical Physics and Education
- Linear Algebra
-
2016.03 - 2022.08 Seoul, South Korea
Awards
- 2023.12.01
NeurIPS 2023 Scholar Award
NeurIPS
Publications
-
2024.09.01 Jump Your Steps: Optimizing Sampling Schedule of Discrete Diffusion Models
Arxiv (under review at ICLR 2025)
-
2024.08.01 -
2024.01.01 Upsample Guidance: Scale Up Diffusion Models without Training
Arxiv (under review)
-
2023.08.01 Geometric Remove-and-Retrain (GOAR): Coordinate-Invariant Explainable AI Assessment
NeurIPS 2023 XAIA Workshop
-
2023.08.01
Skills
| Machine Learning & Generative Models | |
| Diffusion Models | |
| Text-to-Image Models | |
| Safe Generation | |
| Unlearning | |
| Bayesian Modeling |
| Programming | |
| Python | |
| PyTorch | |
| Diffusers | |
| Transformers | |
| R |
Languages
| Korean | |
| Native speaker |
| English | |
| Fluent |
Interests
| Generative AI | |
| Diffusion Models | |
| Safe AI | |
| Explainable AI |
References
| Prof. Junghyo Jo | |
| Associate Professor, Department of Physics Education, Seoul National University (South Korea), jojunghyo@snu.ac.kr |
| Prof. Youngjung Uh | |
| Assistant Professor, Artificial Intelligence Graduate School, Yonsei University (South Korea), yj.uh@yonsei.ac.kr |
| Dr. Chieh-Hsin Lai | |
| Research Scientist, Sony AI; Visiting Assistant Professor, National Yang Ming Chiao Tung University (Taiwan), chieh-hsin.lai@sony.com |
Projects
- 2024.03 - Present
Diffusion Model Safety and Unlearning
Research on enhancing safety of generative models using unlearning algorithms and robust training.
- Safe T2I Generation
- Robust Sampling