About Me

Hi, I’m Jaewon Chu, a 3rd-year M.S.-Ph.D. integrated student in the Department of Computer Science and Engineering at Korea University, conducting research at the Machine Learning and Vision Lab (MLV) under the supervision of Professor Hyunwoo J. Kim.

I’m broadly interested in optimizing generative AI to better serve human needs. These needs are often expressed as black-box functions, such as the performance of API-LLMs (e.g., GPT, Claude, Gemini) or molecule properties. My research focuses on optimizing generative models to produce outputs that meet these human needs, at the intersection of generative models and black-box optimization methods, including Bayesian Optimization, Neural Bandits, and Reinforcement Learning.

📢 I am also actively seeking internship opportunities where I can apply my research skills and contribute to impactful projects.

News

[25.09] Our paper on instruction optimization has been accepted for NeurIPS 2025.
[25.09] I am honored to serve as a reviewer for ICLR 2026.
[25.04] I am honored to serve as a reviewer for NeurIPS 2025.
[25.01] Our paper on latent Bayesian optimization has been accepted for ICLR 2025 (oral).
[24.09] Our paper on latent Bayesian optimization has been accepted for NeurIPS 2024.
[24.02] Our paper on token merging for video transformers has been accepted for CVPR 2024.
[23.09] Our paper on latent Bayesian optimization has been accepted for NeurIPS 2023.
[23.09] Our paper on multi-hop KGQA has been accepted for NeurIPS 2023.
[23.07] Our paper on video question answering has been accepted for ICCV 2023.
[23.03] I started my M.S.-Ph.D. integrated program at MLV Lab.

Selected Publications

(*) denotes equal contribution

• PRESTO: Preimage-Informed Instruction Optimization for Prompting Black-Box LLMs
  Jaewon Chu, Seunghun Lee, Hyunwoo J. Kim
  Advances in Neural Information Processing Systems (NeurIPS), 2025
  
• Inversion-based Latent Bayesian Optimization
  Jaewon Chu^*, Jinyoung Park^*, Seunghun Lee, Hyunwoo J. Kim
  Advances in Neural Information Processing Systems (NeurIPS), 2024
  
• Advancing Bayesian Optimization via Learning Smooth Latent Spaces
  Seunghun Lee^*, Jaewon Chu^*, Sihyeon Kim^*, Juyeon Ko, Hyunwoo J. Kim
  Advances in Neural Information Processing Systems (NeurIPS), 2023