Publications

Peer-Reviewed Publications

Designing physics experiments with artificial intelligenceJonathan Klimesch, Sören Arlt, Carlos Ruiz-Gonzalez, Carla Rodríguez, Xuemei Gu, Philipp Haslinger, Pietro Vischia, Christian Haack, Yehonathan Drori, Rana Adhikari, Markus Arndt, Michael Kagan, Lukas Heinrich, Mario Krenn2 September 2026 | Nature 657, 47–58 (Review)A Review of AI methods for discovering physics experiments, covering hardware search spaces, differentiable simulation, scientific objectives, and algorithms for exploring new experimental designs.

SyConn2: dense synaptic connectivity inference for volume electron microscopyPhilipp J Schubert, Sven Dorkenwald, Michał Januszewski, Jonathan Klimesch, Fabian Svara, Andrei Mancu, Hashir Ahmad, M. Fee, Viren Jain, Joergen Kornfeld24 October 2022 | Nature MethodsSyConn2, an open-source connectome analysis toolkit, is introduced, which works with both on-site high-performance compute environments and rentable cloud computing clusters, and provides a web-based visualization interface that makes these data amenable to complex anatomical and neuronal connectivity queries.

Visualization of laminar-turbulent transition on rotating turbine bladesStefan von Hoesslin, Juergen Gruendmayer, A. Zeisberger, M. Sommer, Jonathan Klimesch, S. Behre, H. Brandies, C. Kähler1 June 2020 | Experiments in FluidsIn this study, Temperature Decline Thermography was enhanced and qualified for use on rotating blades in gas turbines. For the first time, a proof of concept is provided for flow visualization on stationary vanes and rotating blades under realistic flow conditions.

Preprints

Simulating Liquids with Graph NetworksJonathan Klimesch, Philipp Holl, Nils Thuerey14 March 2022 | ArXivThe results indicate that learning models, such as GNNs, fail to learn the exact underlying dynamics unless the training set is devoid of any other problem-specific correlations that could be used as shortcuts.