EfficientTDMPC
Improved MPC objectives for continuous control, reaching 2× the performance of prior methods and significantly outperforming state of the art on the DeepMind Control Suite.
Submitted to NeurIPS 2026.
Thomas Evers
Machine learning researcher working on reinforcement learning, robotics and physical intelligence as a whole.
Research
Resume
Download PDF ↗2023–2026
M.Sc. Electrical Enginerring
Delft University of Technology
Thesis in model-based reinforcement learning, graded 9.5/10.
2023–2024
Founder
Enthusiastech
Built control, computer vision, optimization, and embedded systems for robotics and sensing applications.
2021–2022
Chief Electrical Engineer
Da Vinci Satellite Team, TU Delft
Led six electrical engineers in the design and testing of nanosatellite electronics.
2018–2021
Founder
The Smart Make
Developed a robotics education platform consisting of hardware and software used by more than 100 students.
2020–2023
B.Sc. Electrical Engineering
Delft University of Technology
Electrical engineering, signal processing, embedded systems, and computing.
Skills
Programming
Python, C++, CUDA, PyTorch
Machine learning
Reinforcement learning, VLA and world models, representation learning, model-based control, continuous control, computer vision
Hardware
GPU architecture
Tools
Git, Docker, Linux, LaTeX, MuJoCo, Nsight Systems, Nsight Compute, Slurm, SURF / Snellius
About me
I am lucky enough to see life as the great adventure that it is. My thinking is that the greatest way to experience this adventure is to work hard toward cool goals with great people that want the best for each other and the world. I currently do this in the form of research in machine learning and physical intelligence. It may be the greatest puzzle humanity will ever solve and I am very happy I get to be a small part of it.