Sanjai Rayadurgam

University of Minnesota, University of Minnesota System

Papers

Total Citations

H-Index

About

Sanjai Rayadurgam is a leading researcher in the engineering of safe and reliable autonomous systems, with a primary focus on the rigorous testing and verification of deep neural networks (DNNs). His most cited work, "Input Prioritization for Testing Neural Networks" (2019), addresses the critical challenge of efficiently identifying failure-inducing inputs for DNNs used in safety-critical domains like self-driving cars and medical diagnostics. This contribution is foundational for making AI-driven systems more trustworthy. Rayadurgam also explores the synthesis of systems with random behavior, as seen in his 2020 paper on synthesizing infinite-state systems, where he innovates by moving beyond traditional deterministic Skolem functions to introduce diversity in system behavior. While his citation counts are still growing, reflecting the emerging nature of his field, his work is highly influential in the software engineering and formal methods communities. He is particularly noted for bridging the gap between classical verification and modern machine learning, helping to establish the new discipline of DNN testing.