Felix Aertebjerg

About Me

Recently graduated master student, aspiring PhD sudent. My focus areas are machine learning and probabilistic modeling applied in the sciences, especially in physics and chemistry.

Outside of AI4X, my personal passions are with time-series data and the application of probabilistic models therein. The process of forecasting can be solved in so many interesting ways, and involves much of ML all at once. Particulary, I found GraphCast and Temporal Fusion Transormers very interesting reads.

Research Interests

• Probabilistic Modeling: MCMC, Gaussian Processes, GMMs
• Machine Learning: GNNs, QMC, Diffusion, Bayesian ML
• Comp. Sci.: HPC, MLOps, Python

Notable Projects

• Master Thesis: Enhancing atomistic modeling in Graph Neural Networks through optimized connectivity. GNNs used/improved using our algorithm are primarily MACE and SchNet. More to come soon. :-)
• Project: Variational Quantum Monte Carlo and the Implementation of SPRING in LapNet using JAX. A scaling law of VQMC was also explored.
• GraPE-Chem: Graph-based Property Estimation for Chemistry, a python package on top of PyTorch-Geometric (github, pdf).
• Project: “Uncertainty Estimation for Federated (Distributed) Learning”.
• Bachelor Thesis: “Using AI to study Bifurcation in Dynamical Systems” (pdf).

Last updated: 30/08/2025