Linus Heck

I am a PhD candidate at the Department of Software Science at Radboud University, supervised by Sebastian Junges. My research touches probabilistic model checking, (safe) reinforcement learning, and automated verification, where I am particularly interested in tackling NP-hard problems with new algorithms. I have also worked on dynamical systems and scientific machine learning. Please send me an email!

Publications

Linus Heck, Tim Quatmann, Jip Spel, Joost-Pieter Katoen, Sebastian Junges. "Generalized Parameter Lifting: Finer Abstractions for Parametric Markov Chains." To appear in International Symposium on Automated Technology for Verification and Analysis (ATVA), 2025. [arXiv]

Linus Heck, Maximilian Gelbrecht, Michael T. Schaub, Niklas Boers. "Improving the Noise Estimation of Latent Neural Stochastic Differential Equations." Chaos: An Interdisciplinary Journal of Nonlinear Science, 2025. [Paper (Open Access!)] [arXiv]

Linus Heck, Jip Spel, Sebastian Junges, Joshua Moerman, Joost-Pieter Katoen. "Gradient-Descent for Randomized Controllers under Partial Observability." International Conference on Verification, Model Checking, and Abstract Interpretation (VMCAI), 2022. [Paper] [arXiv]

Software

  • Together with Pim Leerkes and Ivo Melse, I work on Stormvogel, which is a tool for teaching, prototyping, visualizing, and experimenting with probabilistic model checking algorithms. You can try it out online!
  • I'm a frequent contributor to the Storm model checker.

Talks

Thesis Supervision

  • Thomas Maassen: Weighted Model Counting for Bounded Reachability in POMDPs. Master thesis, 2025.

Teaching

  • Model Checking, 2025, Radboud University. Created and graded projects.
  • Automated Reasoning, 2024, Radboud University. Created and graded projects.
  • Formal Systems, Automata, Processes, 2019 and 2020, RWTH Aachen. Held exercise class and graded exercises.
  • Computability and Complexity, 2019, RWTH Aachen. Held exercise class and graded exercises.

Other