معرفی
Matthew Dwyer is a Professor in the Department of Computer Science at the University of Virginia's School of Engineering and Applied Science. With an active research program spanning software engineering, program analysis, and neural network verification, he teaches graduate courses including CS6888 as evidenced by his public course repository.
His research spans traditional software engineering domains and cutting-edge AI verification challenges. Dwyer's work bridges classic program analysis techniques with modern neural network verification, focusing on creating practical tools for ensuring software correctness. His recent publications demonstrate a strategic shift toward addressing verification challenges in deep learning systems while maintaining strong foundations in static analysis and compiler technologies.
Dwyer's publication record shows consistent high-impact contributions with a notable recent emphasis on neural network verification and testing. His work on tools like NeuralSAT and RBT4DNN represents significant advances in making formal verification techniques applicable to modern deep learning systems. The thematic evolution from traditional program analysis to AI safety reflects broader trends in software engineering research.
Dwyer has contributed to major software engineering venues including numerous publications in top conferences and journals. His work on the tipc compiler and associated LLVM passes demonstrates his commitment to practical tool development alongside theoretical advances.
As an educator, Dwyer maintains active engagement with students through graduate course development and research mentorship. His GitHub presence shows commitment to open scholarship with publicly available course materials and research tools.
