Optimization and Trustworthy Machine Learning (OPTML) Group
About
Updated 05/18/26The Optimization and Trustworthy Machine Learning (OPTML) group at Michigan State University is a lab founded by Sijia Liu in the Department of Computer Science and Engineering to advance trustworthy and scalable AI. The group integrates optimization theory with modern deep learning to study topics such as machine learning and deep learning, computer vision, security and signal processing, with a strong emphasis on robust and explainable AI and machine unlearning for large language and vision models. OPTML members publish extensively in top venues like NeurIPS, ICML, ICLR, CVPR and ECCV, and the lab is well funded by federal agencies and industry partners including the National Science Foundation, the U.S. Department of Energy, Open Philanthropy, DSO National Laboratories, Cisco, IBM and Amazon.
Theory of Change
OPTML’s strategy for reducing AI risk is to develop optimization‑driven algorithms that make foundation models robust, explainable and amenable to precise editing. By advancing methods for trustworthy and scalable AI—such as robust optimization for frontier models, stress‑testing for AI safety and machine unlearning to responsibly remove harmful or sensitive knowledge—the group aims to provide practically deployable tools that help developers build safer large‑scale AI systems without sacrificing performance.
Discussion
No comments yet. Be the first to share your thoughts.
Details
- Start Date
- -
- End Date
- -
- Expected Duration
- -
- Funding Raised to Date
- -