Theo Gruner

Theo Gruner joined the Intelligent Autonomous Systems Group as a Ph.D. student in September 2022. Currently, he is focusing on system identification approaches for sim-to-real transfer via likelihood-free inference.

Theo holds a master's degree in Computational Engineering and a bachelor's degree in Applied Mechanics, both from TU Darmstadt. His thesis on "Wasserstein-Optimal Bayesian System Identification for Domain Randomization" was supervised by Fabio Muratore, Boris Belousov, and Jan Peters and won the Freunde-Preis for "best master's thesis at the department of Computer Science."


    •     Bib
      Gruner, T.; Belousov, B.; Muratore, F.; Palenicek, D.; Peters, J. (2023). Pseudo-Likelihood Inference, Advances in Neural Information Processing Systems (NIPS / NeurIPS).
    •     Bib
      Muratore, F.; Gruner, T.; Wiese, F.; Belousov, B.; Gienger, M.; Peters, J. (2021). Neural Posterior Domain Randomization, Conference on Robot Learning (CoRL).
    •     Bib
      Palenicek, D.; Gruner, T.; Schneider, T.; Böhm, A.; Lenz, J.; Pfenning, I. and Krämer, E.; Peters, J. (2024). Learning Tactile Insertion in the Real World, IEEE ICRA 2024 Workshop on Robot Embodiment through Visuo-Tactile Perception.
    •     Bib
      Gruner, T. (2021). Wasserstein-Optimal Bayesian System Identification for Domain Randomization, Master Thesis.

Supervised Projects

Thesis/ProjectStudent(s)TopicTogether with
RL:IP.WS23/24Dennert D., Scherer C., Ahmad F.XXX: eXploring X-Embodiment with RT-XTim Schneider & Daniel Palenicek & Maximilian Tölle
RL:IP.WS23/24Jacobs T.XXX: eXploring X-Embodiment with RT-XTim Schneider & Daniel Palenicek & Maximilian Tölle
RL:IP.WS23/24Böhm A., Pfenning I., Lenz J.Unveiling the Unseen: Tactile Perception and Reinforcement Learning in the Real WorldTim Schneider & Theo Gruner
RL:IP.WS23/24Wang, Y., Li, S.Benchmarking Sequence Models for Discontinuous Dynamical SystemsPuze Liu
RL:IP.SS22Klyushina, A., Rath, M.Black-Box System Identification of the Airhockey TablePuze Liu
RL:IP.SS22Krämer, E.Latent Tactile Representations for Model-based RLTim Schneider & Daniel Palenicek