Joining CISPA as a Postdoctoral Researcher
I joined the Relational Machine Learning group at CISPA Helmholtz Center for Information Security, working on efficient LLMs and multi-agent systems.
Cheeun Hong
Postdoctoral Researcher @ CISPACISPA Helmholtz Center for Information Security
I develop efficient AI methods for reducing the computational cost of deep vision models during training and inference. My work focuses on efficiency optimization, including quantization, pruning, and adaptive computation, with applications in image restoration and video generation. I am particularly interested in controllable and content-adaptive compression methods that preserve model quality while enabling practical deployment on resource-constrained devices.
I am currently a postdoctoral researcher in the Relational Machine Learning group led by Rebekka Burkholz at CISPA, working on efficient LLMs and multi-agent systems through quantization and pruning. I received my Ph.D. from Seoul National University, supervised by Prof. Kyoung Mu Lee.
Keywords Efficient AI · Model Compression · On-device Vision · Generative Models
News
Completed my Ph.D. at Seoul National University
My dissertation received the Distinguished Dissertation Award from Seoul National University.
SneakPeek released on arXiv
Future-guided instructional streaming video generation, from my internship at Meta Superintelligence Labs.
PaperJoining Meta Superintelligence Labs
I will be working as a Research Scientist Intern at Meta Superintelligence Labs (MSL) for 6 months!
Work experience
- Oct 2026 – Present
Postdoctoral Researcher
Relational Machine Learning Group, CISPA Helmholtz Center for Information Security · Germany
Efficient LLMs and multi-agent systems through quantization and pruning.
Supervisor: Rebekka Burkholz
- Jun 2025 – Dec 2025
Research Scientist Intern
Meta Superintelligence Labs (MSL) · Switzerland
Post-training framework for efficient, controllable streaming video generation.
Mentor: Albert Pumarola
- Jun 2019 – Sep 2019
Student Research Intern
Machine Intelligence and Pattern Analysis Lab (MIPAL), Seoul National University · Korea
Designed and optimized a GAN framework for image generation.
Mentor: Prof. Nojun Kwak
- Jun 2018 – Sep 2018
Engineering Intern
SK Hynix, DRAM circuit design team · Korea
Designed an efficient verification system; Grand Prize (1st place) among 100+ interns.
Education
- Mar 2020 – Aug 2026
Integrated Ph.D. in Electrical and Computer Engineering
Seoul National University · Korea
Supervisor: Prof. Kyoung Mu Lee
- Fall 2017
Exchange Student in Computer Science
University of Applied Sciences and Arts Northwestern Switzerland (FHNW) · Switzerland
- Mar 2015 – Feb 2020
B.S. in Electrical and Computer Engineering
Seoul National University · Korea
