I have graduated and am a permanent staff member of the Volkswagen Machine Learning Research Lab.

Philip Becker-Ehmck

Research Interests

Variational Inference, Sequential Latent Variable Models, Intrinsic Motivation, Empowerment, Reinforcement Learning


Publications Google Citations ResearchGate DBLP

Contact Information


Philip Becker-Ehmck joined the Intelligent Autonomous Systems Group as an external PhD student in cooperation with the Volkswagen Machine Learning Research Lab, which is headed by Patrick van der Smagt, in August 2017. Prior to joining the Autonomous Systems Labs, he received his Bachelor and Master of Science degree in computer science from the Technical University of Munich. His Master's Thesis was focused on learning of latent space dynamics via intrinsically motivated exploration. For his PhD, Philip is working on learning more meaningful and hierarchical latent state space representations in particular for physical systems.


    •     Bib
      Becker-Ehmck, P.; Peters, J.; van der Smagt, P. (2019). Switching Linear Dynamics for Variational Bayes Filtering, Proceedings of the International Conference on Machine Learning (ICML).
    •   Bib
      Karl, M.; Becker-Ehmck, P.; Soelch, M.; Benbouzid, D.; van der Smagt, P.; Bayer, J. (2019). Unsupervised Real-Time Control through Variational Empowerment, International Symposium on Robotics Research.