Building the mathematical foundations of reliable and trustworthy machine learning and artificial intelligence systems.
We develop new mathematical tools that help us model and control information in systems that learn from data. We use these tools to delineate fundamental limits of learning, communication, and compression.
Loading theory publications...
We engineer AI and machine learning algorithms that promote privacy, safety, and reliability for all.
Loading AI/ML publications...
We work with a range of applications including AI alignment, privacy, fair machine learning, coded computing, and beyond.
Loading applications publications...