You can find all of my works as preprints on arXiv.
Preprints
Optimal Rates of Convergence for Entropy Regularization in Discounted Markov Decision Processes
Joint work with Semih Çaycı
arXiv preprint, 2024Non-Asymptotic Analysis of Projected Gradient Descent for Physics-Informed Neural Networks
Joint work with Jonas Nießen
arXiv preprint, 2024Embedding Safety into RL: A New Take on Trust Region Methods
Joint work with Nikola Milosevic and Nico Scherf
Accepted at ICML 2025
Publications
11. Fisher–Rao Gradient Flows of Linear Programs and State-Action Natural Policy Gradients
Joint work with Semih Çaycı and Guido Montúfar
SIAM Journal on Optimization, 2024
10. Kronecker-Factored Approximate Curvature for Physics-Informed Neural Networks
Joint work with Felix Dangel and Marius Zeinhofer
NeurIPS 2024
9. Optimization in SciML – A Function Space Perspective
Joint work with Marius Zeinhofer
ICML 2024
8. Geometry and Convergence of Natural Policy Gradient Methods
Joint work with Guido Montúfar
Information Geometry, 2024
7. Achieving High Accuracy with PINNs via Energy Natural Gradients
Joint work with Marius Zeinhofer
ICML 2023
6. Algebraic Optimization of Sequential Decision Problems Joint work with Mareike Dressler, Marina Garotte-Lopéz, Guido Montúfar, and Kemal Rose
Journal of Symbolic Computation, Volume 121, 2022
5. Error Estimates for the Variational Training of Neural Networks with Boundary Penalty
Joint work with Marius Zeinhofer
MSML 2022
4. Notes on Exact Boundary Values in Residual Minimisation
Joint work with Marius Zeinhofer
MSML 2022
3. Uniform Convergence Guarantees for the Deep Ritz Method for Nonlinear Problems
Joint work with Patrick Dondl and Marius Zeinhofer
Advances in Continuous and Discrete Models, 2022
2. Invariance Properties of the Natural Gradient in Overparametrised Systems
Joint work with Jesse van Oostrum and Nihat Ay
Information Geometry, 2022
1. The Geometry of Memoryless Stochastic Policy Optimization in Infinite-Horizon POMDPs
Joint work with Guido Montúfar
ICLR 2022
Extended Abstracts
Solving Infinite-Horizon POMDPs with Memoryless Stochastic Policies in State-Action Space
Joint work with Guido Montúfar
5th Multi-disciplinary Conference on Reinforcement Learning and Decision Making (RLDM 2022)Deep Ritz Revisited
Joint work with Marius Zeinhofer
ICLR Workshop on Integration of Deep Neural Models and Differential Equations, 2020On the Space-Time Expressivity of Residual Networks
ICLR Workshop on Integration of Deep Neural Models and Differential Equations, 2020
🎓 Theses
PhD thesis: Geometry of Optimization in Markov Decision Processes and Neural Network-Based PDE Solvers supervised by Nihat Ay and Guido Montúfar, Download PDF
MSc thesis (University of Warwick): Parameter Estimation of Determinantal Point Processes, supervised by Nikos Zygouras and Theo Damoulas, Download PDF
MSc thesis (University of Freiburg): Universal flow approximation with deep residual networks, supervised by Philipp Harms, Download PDF