Jinsu Kim

MS/PhD Candidate @ Korea University
I am an MS/PhD candidate at Korea University, advised by Prof. Jongheon Jeong. I work on AI safety for generative models—especially large language models and diffusion models—focusing on adversarial robustness, effective use of external knowledge, and mitigating misalignment with human values in rare or unexpected scenarios.
Highlights
🎉 News!
BlurGuard (Poster) Accepted
Our paper "BlurGuard: A Simple Approach for Robustifying Image Protection" has been accepted to NeurIPS 2025!
Selected Publications
BlurGuard: A Simple Approach for Robustifying Image Protection Against AI-Powered Editing Jinsu Kim, Yunhun Nam, Minseon Kim, Sangpil Kim, Jongheon Jeong
NeurIPS 2025 (Poster)
ECoRAG: Evidentiality-guided Compression for Long Context RAG Yeonseok Jeong, Jinsu Kim, Dohyeon Lee, Seung-won Hwang
ACL 2025 Findings
Honors & Awards
Scholar Award (Travel Grant) @ NeurIPS 2025
Projects
BlurGuard

Open-source Python library for robust image protection against AI editing.

Contact
Feel free to reach out for collaborations.
tonmmy222@korea.ac.kr