Lenz Fiedler successfully defended his Ph.D. at the Technische Universität Dresden this week. He is the first graduate to conduct his research at our department. His thesis focused on the development of a machine learning model to replace conventional Density Functional Theory calculations with the goal of accelerating HPC simulations in materials science. During his Ph.D. Lenz became one of the main developers of our software package Materials Learning Algorithms and published several research papers. 

Congratulations Lenz!

Karan's recent work on accelerating electron dynamics simulations using neural networks was accepted as a poster at the ICML workshop "AI for Science: Scaling in AI for Scientific Discovery", which took place on July 26 in Vienna, Austria.

Time-dependent density functional theory (TDDFT) is a common method used to study how electrons behave when exposed to laser fields. In our recent work, we introduced a new way to speed up these simulations using a special type of machine learning model. This model, called an autoregressive neural operator, helps predict the behavior of electrons over time. By incorporating physics-based rules and using detailed training data, our model is more accurate and faster than traditional methods. We tested our approach as a proof of concept on simple toy models and showed that it works well. This method could make it easier to model how molecules and materials respond to lasers in real-time experiments.

Read more about this:

https://doi.org/10.48550/arXiv.2407.09628

Tim Callow and Bartosz Brzoza gave an interview explaining how they came to Görlitz and what they like about working and living in the heart of Europe.

Read more about this:

https://www.welcome-goerlitz-zgorzelec.com/en/blog/detail/82-Research-and-Life-in-Goerlitz

Lenz Fiedler, Bartek Brzoza, and Karan Shah represented us at the Dresden Science Night with a virtual reality installation illustrating concepts from quantum mechanics, materials science, and machine learning.

Read more about this:

https://www.casus.science/events/get-to-know-casus-at-dresden-science-night/ 

We are excited to share that we have received the R&D100 award for our Materials Learning Algorithms package. Congratulations to the team and our collaborators!

Read more about this:

https://www.hzdr.de/db/Cms?pOid=70759&pNid=3438

https://www.casus.science/news/casus-news/prestigious-us-prize-awarded-to-casus-researchers/

Glad to see that our work on artificial intelligence and materials science being featured in HZDR's discovered magazine. You can browse the content below.

Read more about this:

https://www.hzdr.de/HZDR/epaper/entdeckt_1_2023_EN/#0

The arrangement of electrons in matter, known as electronic structure, plays a critical role in fundamental and applied research such as drug design and energy storage. However, the lack of a simulation technique that provides both high fidelity and scalability across different time and length scales has long been an obstacle to the progress of these technologies. We have developed a machine learning method – the Materials Learning Algorithms (MALA) – that replaces traditional electronic structure simulation techniques. MALA enables electronic structure simulations at previously unattainable length scales.

Read more about this:

https://www.hzdr.de/db/Cms?pOid=69680&pNid=3438

https://www.casus.science/news/casus-news/machine-learning-takes-materials-modeling-into-new-era/

Our second hackathon aimed to refine the Materials Learning Algorithms (MALA) into a production-grade code, enabling efficient computations on both central processing units (CPU) and graphical processing units (GPU).

Read more about this:

https://www.casus.science/news/casus-news/casus-propels-the-development-of-the-mala-framework/

Congratulations, Karan, for being accepted into the program for the second time! 

Given the growing demand for data science skills, many physics programs are beginning to incorporate data science into their curriculums. A team of physicists led by the Topical Group on Data Science (GDS) and funded by the American Physical Society (APS) is therefore developing teaching modules for data science that can be easily integrated into undergraduate physics courses. The project, known as the Data Science Education Community of Practice (DSECOP), was designed to bridge the gap between traditional physics education and the skills physicists need in industry. 

Read more about this:

https://www.casus.science/news/casus-news/casus-scientist-reselected-for-dsecop-fellowship/

Congratulations, Debanjan, for receiving the Fulbright-Nehru fellowship! 

This fellowship, supporting early-career Indian researchers, offers the opportunity for a long-term stay at a top US university. Debanjan was selected through a highly competitive process among applicants from a variety of scientific disciplines.

Starting in March 2023, the fellowship will fund Debanjan's research for the next two years at Purdue University (USA). His focus will be on quantum machine learning leveraging quantum computing concepts to design more accurate and efficient machine learning algorithms. 

Read more about this:

https://www.casus.science/news/casus-news/fulbright-nehru-stipendium-geht-an-debanjan-konar/

We are excited to join the outreach activities "oha! Abenteuer Wissenschaft" in Görlitz with a virtual reality installation showcasing our work on accelerating atomistic simulations with machine learning to the broad public.

Read more about this:

https://www.casus.science/news/casus-news/oha-abenteuer-wissenschaft/

Congratulations, Karan, for being selected as a fellow of the Data Science Education Community of Practice (DSECOP) program of the American Physical Society (APS). The DSECOP program aims at developing teaching materials that integrate artificial intelligence and machine learning into the undergraduate physics curriculum. Instead of a separate course for machine learning, the modules will be focused on data science applications in the context of physics problems and be taught as a part of existing standard undergraduate physics courses. 

Read more about this:

https://www.casus.science/news/casus-news/casus-scientist-joins-international-team-to-develop-ai-teaching-material/

We received funding from the Helmholtz Association's Artificial Intelligence Cooperation Unit (Helmholtz AI) for our project Machine-learning based synthetic data generation for rapid physics modeling (SynRap).  In collaboration with the Deutsches Elektronen-Synchrotron (DESY, Hamburg), we will investigate whether data on the behavior of physical systems can be generated more quickly using neural networks in the form of so-called synthetic data. Our project was selected for funding through a competitive process. The fifteen winning teams will receive a total of 6.2 million euros over the next few years to carry out their projects.

Read more about this:

https://www.hzdr.de/db/Cms?pOid=63571&pNid=0&pLang=en

https://www.casus.science/news/casus-news/seeking-a-faster-pathway-to-synthetic-data/