About

I am an M.S. student in the School of Computing at the Korea Advanced Institute of Science & Technology (KAIST), advised by Prof. Hyunwoo J. Kim in the MLV Lab.

My research interests lie in building AI agents that can automate complex, multi-step tasks, and in advancing multimodal AI systems that better understand, reason over, and interact with visual information, particularly videos.

Recently, I have been particularly interested in the following research directions:

Agent systems for video understanding and reasoning, particularly for long-form and egocentric videos, as well as methods for efficient video inference on edge devices.
Prompt and agent optimization methods, with an emphasis on improving their computational and data efficiency.
Post-training methods for Vision-Language Models (VLMs) to improve multimodal understanding and reasoning.

I am open to research collaborations and internship opportunities. Please feel free to reach out via email.

News

Jul. 2026Our paper VideoSearch-R1: Iterative Video Retrieval and Reasoning via Soft Query Refinement has been accepted to ECCV 2026!
Feb. 2026I received my B.S. in the School of Computing at KAIST.
Feb. 2026Our paper MoE-GRPO: Optimizing Mixture-of-Experts via Reinforcement Learning in Vision-Language Models has been accepted to CVPR 2026!

Education

Korea Advanced Institute of Science & Technology (KAIST), School of Computing - M.S. Student, Sep. 2026 - Present
Korea Advanced Institute of Science & Technology (KAIST), School of Computing - B.S., Mar. 2021 - Feb. 2026