Cognitive Psychology

Neuroscience

Linguistics

Philosophy

Computer Science

Cognitive Science Connections / Faculty Research Spotlight

How can we design AI systems that follow specific rules, respect physical constraints, and produce safe and reliable results? This question is central to the work of Ferdinando (Nando) Fioretto, Associate Professor of Computer Science at the University of Virginia and director of the Responsible AI for Science and Engineering (RAISE) research group.

Professor Fioretto and his team study how to incorporate physical laws, mathematical constraints, and safety requirements into generative AI models. Their research addresses problems in areas such as robotics, scientific discovery, language generation, and software development. One challenge they are particularly interested in is how to make AI models follow rules they were never explicitly trained to follow.

In this edition of Cognitive Science Connections, we speak with Professor Fioretto about his research, including how AI models can generate and revise their outputs while respecting constraints, and how these processes compare with human problem-solving. We also discuss the challenges that remain in developing trustworthy AI systems and how undergraduate students can get involved in his research through coursework and opportunities in his lab.

The list of approved courses for the 2027 Spring and J-Terms has been updated! Make sure you are reviewing your coursework eligibility before you enroll!

Have Questions? Start by viewing our FAQ page just for transfer students found here