Generative Intelligence & Discovery Laboratory
About Us
Our lab works at the frontier of generative AI, developing models that faithfully capture the full complexity of real-world data. On the fundamental side, we improve their fidelity and coverage while advancing inclusiveness, trustworthiness, and reliability. On the application side, we extend this work to science and technology, from material design to drug response prediction, and to physical AI, connecting generative intelligence to real-world sensing and diagnostics.
Generative AI
Exploring generative models, including diffusion models and LLMs, and how they generate, represent, and reason about data.
Trustworthy AI
Improving the inclusiveness, fairness, safety, and privacy of generative models.
AI for Science
Applying generative modeling to science, including material design and drug response prediction.
Physical AI
Extending generative approaches to physical systems, including sensor fusion and real-world diagnostics.
Latest News
MotionCFG accepted to NeurIPS 2026
Prof. Um will serve as an Area Chair for ICLR 2027.
Prof. Um delivered an invited talk on AI for Science at the NRF/DFG Joint Workshop.
Prof. Um joined DGIST as an Assistant Professor in the Department of EECS.
We Are Recruiting Students & Interns
We welcome highly motivated M.S./Ph.D. students and undergraduate interns. If you are interested in our lab, please send your CV, transcript, and research interests to Prof. Um.
Contact Information