The AMC Seminar is back! After spring break, the summer semester seminar series on Atomistic Modeling-related topics will start next week. Day and time changed to Wednesdays, 10 am, location is still the MIBE lecture hall.

Kicking off our Summer Term Seminar Series, we’re thrilled to welcome Dr. Jonathan Schmidt as our first speaker. ETH Zürich Researcher Schmid has been awarded an SNSF Ambizione Grant for his project “Universal Machine Learning Interatomic Potentials for Magnetic Systems”, and since January 2026, he has been conducting the research at the Laboratory of Computational Science and Modelling (COSMO) at EPFL Lausanne, Switzerland. The project aims to develop a universal machine-learning-based interatomic potential capable of accurately describing magnetic materials — a key challenge in materials modeling. 

Large-scale ML models now enable materials scientists to explore and relax millions of crystal structures at unprecedented speed, exemplified by the Alexandria database, which vastly expands the catalog of stable 1D–3D materials and supports the creation of universal ML force fields. In his talk on “Machine Learning Materials Modelling”, Schmidt will explain how these advances, along with ML‑corrected DFT methods and ML‑accelerated simulations, are closing the gap between ab initio theory and real experimental conditions, reshaping what problems can be tackled in modern materials science.

Date: Wednesday, March 18, 2026, 10 am

Location: MIBE Lecture Hall

Abstract:

Large-scale machine learning (ML) models are transforming materials modelling by dramatically increasing the scale and speed at which materials can be explored. A central example of this development is the Alexandria database, which we created through high-throughput density functional theory (DFT) searches accelerated by crystal-graph-attention networks. With millions of DFT-relaxed crystal structures spanning one-, two-, and three-dimensional materials, Alexandria multiplies the number of known stable materials and provides a foundation for training next-generation universal ML force fields. 

The talk will further explore how ML methods help bridge the gap between ab initio simulations and real experimental conditions. Examples include interpretable ML models that correct systematic errors in density functional theory and ML force-field–accelerated simulations that enable the study of materials questions previously beyond computational reach. Together, these advances highlight the transformative impact of ML on modern materials science.