Automated reasoning, formal methods, computer-aided programming, logic, game theory
Swarat Chaudhuri is a formal methods (FM) researcher with a particular interest in the intersection of FM and artificial intelligence. Much of his recent work is on program synthesis, the problem of automatically discovering high-level programs that fit a dataset of example behaviors and also satisfy a set of logical correctness criteria. This problem can be seen as a form of machine learning in which the learned function is represented programmatically and comes with some formal guarantees. In his view, progress on this problem can bring us closer to the goal of reliable and interpretable artificial intelligence, and has radical implications for how we think about software.