Moritz Schneider

Hi! I am a Senior AI Engineer at Agile Robots in Munich scaling reinforcement learning with world-action models. In parallel, I am a PhD student at the Neurorobotics Lab at the University of Freiburg, previously in collaboration with Bosch Corporate Research, focusing on pre-training approaches for reinforcement learning agents. I am advised by Prof. Joschka Boedecker. Before starting my PhD, I earned a Master of Science in Automation Engineering and a Bachelor of Science in Business Administration and Mechanical Engineering from RWTH Aachen University. During this time, I worked on disentangled representation learning and interpretability methods for reinforcement learning.

Research

My research aims to uncover the principles required to build general embodied AI. I focus on pre-training approaches that let agents learn from diverse data before seeing any task-specific data. My PhD work explores unsupervised objectives like empowerment and learning latent actions from videos without action labels, so that agents and their world models acquire a general notion of action that can be fine-tuned efficiently for downstream tasks. At Agile Robots, I bring these ideas to real robots by combining world-action models with reinforcement learning.

Publications

• Information-Theoretic Policy Pre-Training with Empowerment
  arXiv, 2025, Moritz Schneider, Robert Krug, Narunas Vaskevicius, Luigi Palmieri, Michael Volpp, Joschka Boedecker
• The Surprising Ineffectiveness of Pre-Trained Visual Representations for Model-Based Reinforcement Learning
  NeurIPS, 2024, Moritz Schneider, Robert Krug, Narunas Vaskevicius, Luigi Palmieri, Joschka Boedecker
• Interpretable Domain Randomization for Reinforcement Learning with Disentangled Representations
  Master Thesis, 2021, Moritz Schneider
• How Do You Act? An Empirical Study to Understand Behavior of Deep Reinforcement Learning Agents
  arXiv, 2020, Richard Meyes, Moritz Schneider, Tobias Meisen