Binxu Wang

Binxu Wang is an Inaugural NeuroAI Research Fellow at the Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University, starting in Fall 2023. He holds a Ph.D. in Neuroscience from Harvard Medical School, advised by Dr. Carlos Ponce, after completing his doctoral studies at Washington University in St. Louis. His research focuses on using optimization, generative models, and geometry to bridge visual information processing in the primate brain and deep neural networks. Key interests include analyzing diffusion models' spectral bias, the geometry of image manifolds, and neural dynamics in ventral stream alignment.

He has received the prestigious NeuroAI Research Fellowship at the Kempner Institute. His work spans theoretical insights into neural representations, practical applications of generative models, and interdisciplinary approaches linking neuroscience and AI. Notable contributions include uncovering hidden linear structures in score-based models and exploring the geometry of natural image manifolds.

Wang has presented his research at venues like the Cognitive & Neural Computation Lab (Yale), NAISYS2024, and the KITP DL2023 Program. His research emphasizes a closed-loop approach, integrating computational theory with experimental neuroscience to advance understanding of brain function.