Song Mei

Song Mei is an Assistant Professor at the University of California, Berkeley, jointly appointed in the Department of Statistics and the Department of Electrical Engineering and Computer Sciences. Her research bridges statistics, machine learning, information theory, computer science, and statistical physics, focusing on theoretical foundations of AI, including language models, diffusion models, quantum algorithms, and uncertainty quantification.

• PhD in Computational and Mathematical Engineering, Stanford University (2020)
• BS in Mathematics, Peking University (2014)

Her publications span topics like language models, diffusion models, deep learning theory, reinforcement learning, high-dimensional statistics, and quantum algorithms. She has received prestigious awards such as the Sloan Research Fellowship and NSF CAREER Award. Notably, she is on leave at OpenAI (May 2025), indicating ongoing industrial collaboration.

Scientific Awards and Recognitions:

• Sloan Research Fellow (2025)
• NSF CAREER Award (2025, 2024)
• Okawa Research Grant (2024)
• Amazon Research Award (2024)
• Google Faculty Research Award (2024)

Her research group includes PhD students from statistics, EECS, and mathematics backgrounds, with former mentees now at institutions like OpenAI, UCLA, and Flatiron Institute. Song Mei contributes to theoretical advancements in AI, focusing on model generalization, optimization, and alignment safety.