About Me.
I am a graduate student at KAIST, where I am fortunate to be advised by Professor Seunghoon Hong. My research interests lie in the advancement of Generative Models, with a specific focus on Flow Matching and Diffusion.
Education
- M.S. in Computer Science, KAIST, South Korea, 2025–Present
- B.S. in Computer Science and Electronic Engineering (Double Major), KAIST, South Korea, 2021–2025
Publications
Conference Papers
FlowBind: Efficient Any-to-Any Generation with Bidirectional Flows
Published in ICLR, 2026
Efficient any-to-any multimodal generation with bidirectional flows.
Recommended citation: Yeonwoo Cha*, Semin Kim*, Jinhyeon Kwon, Seunghoon Hong. (2026). "FlowBind: Efficient Any-to-Any Generation with Bidirectional Flows." ICLR.
Download Paper
Simulation-Free Training of Neural ODEs on Paired Data
Published in NeurIPS, 2024
Training neural ODEs on paired data without simulation.
Recommended citation: Semin Kim, Jaehoon Yoo, Jinwoo Kim, Yeonwoo Cha, Saehoon Kim, Seunghoon Hong. (2024). "Simulation-Free Training of Neural ODEs on Paired Data." NeurIPS.
Download Paper
Preprints / Workshop Papers
Reward-Agnostic Prompt Optimization for Text-to-Image Diffusion Models
Published in arXiv, 2025
Prompt optimization for text-to-image diffusion models without requiring a task-specific reward.
Recommended citation: Semin Kim, Yeonwoo Cha, Jaehoon Yoo, Seunghoon Hong. (2025). "Reward-Agnostic Prompt Optimization for Text-to-Image Diffusion Models." arXiv.
Download Paper