Deokyeong Lee

Hi! I’m Deokyeong Lee, currently a Research Intern at the Visual & Geometric Intelligence Lab at Seoul National University, advised by Prof. Jaesik Park.

My research interests lie in the theoretical foundations of generative models, with a particular focus on Diffusion Models and Flow-based models. I am deeply interested in exploring how these frameworks can be generalized and applied across diverse modalities, including text and image generation.

Prior to my current focus, I conducted research on the efficiency and reliability of foundation models. My previous work includes accelerating LLM inference through hardware-aware algorithms and mitigating hallucinations in Multi-modal LLMs. These experiences drive my current passion for developing fundamentally unified and efficient generative algorithms.

For more details, please refer to my Curriculum Vitae.