Yu Lei (雷宇) is a first-year Ph.D. student at the Visual Information Processing and Learning (VIPL) Lab, Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS), where he explores AIGC, video generation, world models, and preference learning. His current research focuses on personalized image generation, preference optimization for visual generative models, and physically consistent world models.

Before joining ICT, Yu obtained his B.Eng. in Artificial Intelligence from Sichuan University (June 2025), where he began his journey in weakly supervised learning for medical imaging. As a research intern at TeleAI, China Telecom, he worked on personalized safety alignment for text-to-image diffusion models, preference optimization for visual generative models, and world-model research under the mentorship of Dr. Jinbin Bai and Dr. Kaidong Yu.

"I believe AI should understand what humans truly want, not just what they say they want."

Education

Institute of Computing Technology (Sep. 2025 – Now)

Ph.D. in Computer Science and Technology

Advisor: Prof. Liang Li

Sichuan University (Sep. 2021 – Jun. 2025)

B.Eng in Artificial Intelligence

Advisor: Prof. Lituan Wang