Patrick Gelß

Patrick Gelß is a Postdoctoral Researcher at Zuse Institute Berlin, working in the AI in Society, Science, and Technology department. He leads the iol.QUANT research group focused on quantum-inspired and tensor-based methods, with affiliations to Freie Universität Berlin and collaborative projects like MATH+ Cluster of Excellence.

• Diploma in Mathematics/Physics (2013, Freie Universität Berlin)
• PhD in Mathematics (2017, summa cum laude, Freie Universität Berlin)
• Postdoctoral roles at Zuse Institute Berlin (2020-present) and CRC 1114 (2017-2022)

His research bridges tensor decompositions, quantum computing, and dynamical systems, with notable contributions to:

• Graph Isomorphism: Continuous optimization approaches via doubly stochastic matrices and Frank-Wolfe algorithms
• Quantum Simulation: Tensor network representations for quantum circuits and the open-source WaveTrain package
• KvN Mechanics: Mathematical analysis of existence/uniqueness solutions for bounded domains
• Fredholm Networks: Novel training paradigms for neural networks using integral equations

Scientific achievements include:

• Developing Scikit-TT for tensor-train computations
• Organizing major workshops (QML@SC2024, TMQS 2024)
• Leading the Thematic Einstein Semester 2024 on Mathematics for Quantum Technologies

His work spans chemical kinetics, quantum dynamics, and quantum machine learning, with applications to CO oxidation models, exciton-phonon systems, and mutational hierarchies in medicine.