Zhuang Liu

Zhuang Liu is an Assistant Professor of Computer Science at Princeton University, where he leads a research group focused on deep learning and computer vision. His work spans vision and language, unified by a focus on deep learning methods, representations, and architectures. Prior to joining Princeton, he was a Research Scientist at Meta AI Research (FAIR) in New York City. He received his Ph.D. from UC Berkeley and his B.E. from Tsinghua University, both in Computer Science.

His educational background includes:

• Ph.D. in Computer Science, University of California, Berkeley, 2022
• B.E. in Computer Science, Tsinghua University (Yao Class)

Liu's research focuses on empirical approaches to understanding how deep learning models work and behave. He explores simple approaches to gain empirical insights into neural networks, often challenging existing beliefs in architectures, training, pruning, and datasets. His work spans multiple domains including computer vision, natural language processing, and multimodal learning. He has made significant contributions to the field, including DenseNet and ConvNeXt architectures, which have influenced modern neural network design.

His recent publications reveal a strong focus on understanding and improving large language models and vision-language systems. His work examines idiosyncrasies in LLMs, pruning approaches for efficient inference, visual shortcomings of multimodal systems, and bias in large-scale visual datasets. He also investigates fundamental questions about neural network architectures, exploring the relationship between ConvNets and Transformers, and developing normalization-free transformer variants. His research consistently bridges theoretical insights with practical applications, as evidenced by his numerous conference publications and industry collaborations.

His notable scientific achievements include:

• CVPR Best Paper Award
• NeurIPS'18 Compact Neural Networks Workshop Best Paper Award

Professor Liu actively mentors students and postdoctoral researchers, currently advising seven Ph.D. students including Wenhao Chai, Tony Chen, Sachin Konan, Taiming Lu, Zhuorui Ye, David Yin, and Boya Zeng. He has successfully guided graduates such as Jiachen Zhu (now at Skild AI) and Mingjie Sun (now at Thinking Machines Lab). His research has practical implications for improving the efficiency, interpretability, and fairness of deep learning systems, with applications across multiple domains.

His research group maintains active collaborations with industry partners including Meta FAIR, NVIDIA Research, and Adobe, and regularly hosts research interns. The group's work has established Zhuang Liu as a leading voice in empirical deep learning research, with invitations to speak at top academic institutions including Harvard, Stanford, Columbia, and Boston University.