 |
Zhihui Zhu
|
We are seeking students interested in advancing the foundations of Machine Learning and Quantum Information. In ML, we investigate the foundations of learning and modeling to build intelligent, trustworthy, and adaptable systems, with research spanning world models, generative AI, interpretability, and embodied intelligence. In quantum, we develop learning and computational methods to understand, design, and harness quantum systems for scientific discovery and computation. We welcome candidates from diverse backgrounds, including computer science, electrical engineering, mathematics, and physics.
Recent Selected Publications:
Learning, Modeling, and Acting in the World
J. Jiang, T. Ding, Z. Zhu, ‘‘DeltaEvolve: Accelerating Scientific Discovery through Momentum-Driven Evolution", International Conference in Machine Learning (ICML), 2026.
J. Jiang, H. Zhu, Z. Zhu, ‘‘SMCEvolve: Principled Scientific Discovery via Sequential Monte Carlo Evolution", 2026.
Representation and Interpretability
X. Zhu, M. Mahdi, Z. Zhu, ‘‘AbsTopK: Rethinking Sparse Autoencoders For Bidirectional Features", International Conference on Learning Representations (ICLR), 2026.
Y. Dong, J. Jiang, Z. Zhu, Xia Ning, ‘‘Understanding Task Vectors in In-Context Learning: Emergence, Functionality, and Limitations," International Conference on Learning Representations (ICLR), 2026.
J. Jiang, J. Zhou, Z. Zhu, ‘‘Tracing Representation Progression: Analyzing and Enhancing Layer-Wise Similarity," International Conference on Learning Representations (ICLR), 2025.
Quantum Information
Z. Qin, J. M. Lukens, B. T. Kirby, and Z. Zhu, ‘‘Enhancing Quantum State Reconstruction with Structured Classical Shadows," npj Quantum Information, 2025.
Z. Qin, C. Jameson, Z. Gong, M. B. Wakin, and Z. Zhu, ‘‘Quantum State Tomography for Matrix Product Density Operators," IEEE Transactions on Information Theory, vol. 70, pp. 5030 - 5056, 2024.
Z. Zhu, J. Lukens, and B. Kirby, ‘‘On the connection between least squares, regularization, and classical shadows," Quantum, vol. 8, 2024.
News:
[Jan 2026] 3 papers at ICLR 2026 on interpretability and in-context learning, 1 paper (Oral) at CPAL 2026 on model collapse, 1 paper at AISTAS 2026 on faithfully evaluating machine unlearning through internal representation, and journal papers at IEEE TQE and Physical Review A on quantum information.
[Nov 2025] Congratulations to Alina Liu and Chris Hill (high school interns from Columbus Academy) and Jiachen Jiang for receiving the Best Paper Award at ICDM UGHS Symposium 2025 for the work ‘‘Can Reasoning LLMs Eliminate Conformity in Multi-Agent Systems?".
[Oct 2025] Gave a tutorial lecture on “Emergence of Low-dimensional Representations in Deep Models” at ICCV 2025 and a ECE seminar on ‘‘Efficiently Probing Quantum Systems with Compressive Measurements" at University of Hawaii at Manoa.
[Aug 2025] Zhen defended his PhD thesis and will join University of Michigan as MICDE Postdoctoral Fellow. His PhD work received graduate research award from CSE of OSU.
[June 2024] Gave a tutorial on “Learning Deep Low-dimensional Models from High-Dimensional Data: From Theory to Practice” at CVPR 2024.
[May 2024] Thrilled to receive the ORAU Ralph E. Powe Junior Faculty Enhancement Award.
[April 2024] Tutorial on “Understanding Deep Representation Learning via Neural Collapse” at ICASSP’24 with Prof. Laura Balzano, Dr. Peng Wang, and Prof. Qing Qu. Slides: Lecture-1, Lecture-2
[Jun 2023] Gave a short course on ‘‘Learning Nonlinear and Deep Low-Dimensional Representations from High-Dimensional Data: From Theory to Practice" at ICASSP 2023.
[Jan 2023] Co-organized and gave a tutorail at the 3rd SLowDNN Workshop at MBZUAI, Abu Dhabi, MBZUAI.
[Dec 2022] Received CQISE Partnership Seed Award (PSA) and will collaborate with Brian Kirby (ARL) on quantum network.
[May 2022] Gave a 10-hours short course (with Sam Buchanan, Yi Ma, Qing Qu, John Wright, and Yuqian Zhang) on ‘‘Low-Dimensional Models for High-Dimensional Data: From Linear to Nonlinear, Convex to Nonconvex, and Shallow to Deep" at IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).
[Nov 2021] Co-organizing the 2nd workshop on ‘‘Seeking Low-dimensionality in Deep Neural Networks (SLowDNN)’’, Nov. 22nd – Nov. 23rd, 2021.
[Nov 2020] Co-organizing IEEE workshop on ‘‘Seeking Low-dimensionality in Deep Neural Networks (SLowDNN)’’, Nov. 23rd – Nov. 24th, 2020.