👨🏻‍💻 I’m a Research Scientist at Sakana AI.

🔬 My research interests are in Openended Learning using large foundation models!

📽️ I am also interested in assisting others on their path in the world of Machine Learning and academia.

Selected Experience

👨🏻‍🔬 Professional Experience

Currently, I am working as a Research Scientist at Sakana AI exploring how to develop an openended algorithm to coevolve a population of agents and the environment they are in.
Before this, I explored latent reasoning by combining diffusion models with large language models and how to train a process reward model for training multimodal large language models as UI-Agents.

During my final year at college I was working as a Student Researcher at the Reliable Multimodal AI Lab at the TU Darmstadt under the supervision of Prof. Marcus Rohrbach working on Generative Video-Language Modelling.

Additionally, I have professional experience working as an Applied Machine Learning Student Researcher at the Fraunhofer Heinrich Hertz Institute.
There I worked on dynamic traffic flow forecasting using Graph Neural Networks.

🤖 Open Source Contributions

I have experience contributing to Arena Bench a large open-source project for robotic obstacle avoidance using Deep Reinforcement Learning.

Moreover, I have published a respective paper at the IROS conference and in the Robotics and Automation Letters (RA-L) journal.

📜 Reimplementing and Reproducing Papers

I have experience with independent research. I have implemented the Reward Constrained Policy Optimization paper into stable-baselines3 PPO and reproduced the original results by running and tracking experiments.

To accompany this work, I have submitted a blog post to the ICLR Blogposts Track communicating the paper’s theory and my results.

Feel free to look at my specific portfolio entry.

📚 Teaching and Community Contributions

To further contribute to the Machine Learning community, I have a YouTube and Medium channel where I publish educational Machine Learning content. Moreover, I try my best to be a reviewer at top conferences that people don’t hate. I even got an outstanding reviewer award at the CVPR 2025 conference!