Faculty DirectoryHan Liu
Orrington Lunt Professor of Computer Science and (by courtesy) Industrial Engineering & Management Sciences
Professor of Statistics
Contact
2233 Tech DriveMudd Room 3119
Evanston, IL 60208-3109
Email Han Liu
Website
Departments
Education
Ph.D. Machine Learning and Statistics, Carnegie Mellon University, Pittsburgh, PA
Biography
Han Liu develops the foundations and systems of scientific intelligence, connecting reasoning agents, scientific world models, and autonomous laboratories. He is the Orrington Lunt Professor of Computer Science and Professor of Statistics at Northwestern University, where he directs MAGICS Lab and the Center for Foundation Models and Generative AI. His research asks how AI can turn scientific experience into better models of the world and better decisions about what to investigate next.
With students and collaborators, Liu has developed genome foundation models including DNABERT, DNABERT-2, and GenomeOcean, and contributed to SciSciGPT, a research-agent collaboration published in Nature Computational Science. He is a co-principal investigator of DREAM, an NSF-supported $20 million initiative developing a national cloud laboratory for protein engineering. Together, these efforts connect learning and reasoning with experimental evidence.
His work builds on a foundation in statistical machine learning, high-dimensional inference, graphical models, and optimization. Before Northwestern, he held faculty appointments at Princeton University and Johns Hopkins University and directed the deep reinforcement learning center at Tencent AI Lab. He earned a joint PhD in Machine Learning and Statistics from Carnegie Mellon University. His honors include the Presidential Early Career Award for Scientists and Engineers and an Alfred P. Sloan Research Fellowship in Mathematics.
Research Interests
Scientific Intelligence: From Discovery to Invention
Han Liu develops the foundations and systems of scientific intelligence, connecting self-improving scientific agents, scientific world models, and autonomous laboratories. His central question is how each research cycle can improve both what AI understands and what it can help us build.
Discovery asks how the world works. Invention asks what we can build within it. His research connects these goals through a shared cycle: reason, model, design, experiment, verify, and learn. The agenda spans three connected frontiers:
- Self-improving Scientific Agents. Liu studies memory, reasoning, adaptation, and tool use: how agents can learn from evaluated outcomes and make better research decisions. He is a coauthor of SciSciGPT, a collaboration with Northwestern’s Center for Science of Science and Innovation published in Nature Computational Science, which connects scholarly data and research tools through a conversational workflow.
- Scientific World Models. Liu develops representations of biological and physical systems for prediction, uncertainty quantification, and intervention design. His work with students and collaborators includes the open-source genome foundation models DNABERT and DNABERT-2, and GenomeOcean, developed with the DOE Joint Genome Institute and collaborators. These projects provide reusable foundations for genomic representation learning and sequence generation.
- Autonomous Laboratories. Liu investigates how model-guided design, simulation, automation, and experimental feedback can work together. He serves as Co-PI and AI lead of DREAM, an NSF-supported $20 million initiative developing a national cloud laboratory for protein engineering. His group’s PEV work studies how measured protein outcomes can guide subsequent design decisions.
The long-term aim is to connect these advances into systems that learn across discovery and engineering cycles. Progress is measured through better predictions, more informative experiments, and more effective designs.
Grounded in statistical machine learning, high-dimensional inference, graphical models, and optimization, Liu’s research links mathematical foundations with open-source systems and experimental infrastructure. He welcomes prospective students, scientific collaborators, and industry partners interested in advancing this agenda.
Research program | Publications | Open-source systems
Selected Publications
- Erzhuo Shao, Yifang Wang, Yifan Qian, Zhenyu Pan, Han Liu, Dashun Wang. SciSciGPT: advancing human–AI collaboration in the science of science. Nature Computational Science 6: 301–315 (2026).
- Jerry Yao-Chieh Hu, Hude Liu, Jennifer Yuntong Zhang, Han Liu. In-Context Algorithm Emulation in Fixed-Weight Transformers. International Conference on Learning Representations (ICLR) (2026).
- Yibo Wen, Chenwei Xu, Jerry Yao-Chieh Hu, Kaize Ding, Han Liu. Pareto-Optimal Energy Alignment for Designing Nature-Like Antibodies. Advances in Neural Information Processing Systems 38 (2025).
- Weimin Wu, Teng-Yun Hsiao, Jerry Yao-Chieh Hu, Wenxin Zhang, Han Liu. In-Context Learning as Conditioned Associative Memory Retrieval. Proceedings of the 42nd International Conference on Machine Learning 267: 67300–67325 (2025).
- Zhihan Zhou, Yanrong Ji, Weijian Li, Pratik Dutta, Ramana Davuluri, Han Liu. DNABERT-2: Efficient Foundation Model and Benchmark For Multi-Species Genomes. International Conference on Learning Representations (ICLR) (2024).
- Yanrong Ji, Zhihan Zhou, Han Liu, Ramana V. Davuluri. DNABERT: pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genome. Bioinformatics 37(15): 2112–2120 (2021).
- Han Liu, John Lafferty, Larry Wasserman. The Nonparanormal: Semiparametric Estimation of High Dimensional Undirected Graphs. Journal of Machine Learning Research 10: 2295–2328 (2009).