Rate My Professor David Sprunger

Sprunger's research specializations lie at the intersection of coalgebra, logic, and category theory, with interests in mathematical foundations for machine learning and neural networks, quantitative bisimulations, and formal methods for cyber-physical systems. His key publications include 'The differential calculus of causal functions' (with Bart Jacobs, 2019), 'Differentiable Causal Computations via Delayed Trace' (with Shin-ya Katsumata, LICS 2019), 'Relational Differential Dynamic Logic' (with Jérémy Dubut, Shin-ya Katsumata, Ichiro Hasuo, Juraj Kolčák, and Akihisa Yamada, 2019), 'Quantitative bisimulations using coreflections and open morphisms' (with Jérémy Dubut, Shin-ya Katsumata, and Ichiro Hasuo, 2018), 'Fibrational Bisimulations and Quantitative Reasoning' (with Jérémy Dubut, Shin-ya Katsumata, and Ichiro Hasuo, CMCS 2018), 'Neural Nets via Forward State Transformation and Backward Loss Transformation' (with Bart Jacobs, 2018), 'Precongruences and parametrized coinduction for logics for behavioral equivalence' (with Lawrence Moss, CALCO 2017), 'A complete logic for behavioural equivalence in coalgebras of finitary set functors' (CMCS 2016), and 'Eigenvalues and Transduction of Morphic Sequences' (with Lawrence Moss, William Tune, and Jörg Endrullis, DLT 2014). Additional works cover linearization of automatic arrays, weave specifications, and variadic sequences (MFPS 2013).