As part of its efforts to advance science through the use of artificial intelligence, the U.S. Department of Energy recently announced an investment of more than $320 million in projects and awards to accelerate AI capabilities.
These investments will support the Genesis mission, a broad effort by the U.S. Department of Energy to develop an integrated platform that connects the world's best supercomputers, laboratory facilities, AI systems, and unique datasets across every major scientific discipline.
Professor Julia R. Gell of the Department of Statistics is a key member of one of these new projects, the LEearning-Accelerated Domain Science (LEADS) Institute. The institute aims to make scientific machine learning available to domain scientists.
Research at the LEADS Institute will contribute to the development of new algorithms for accurate and efficient exploration of large-scale, complex data and real-time information extraction using digital twin-assisted optimal control.
Gel's own research interests focus on statistical topological and geometric algorithms, with a particular focus on efficient graph learning, graph-based AI, and associated uncertainty quantification.
“I am excited to contribute to the LEADS initiative because it breaks traditional disciplinary boundaries and brings statistical and mathematical foundations to the forefront of innovation in scientific computing,” said Gell, who joined the Virginia Tech faculty in 2024 after serving as a program officer at the National Science Foundation. “We also see this as an opportunity to redefine and emphasize the unique role that modern statistical science plays, not just in scientific machine learning, but in the advancement of AI in general.”
LEADS has researchers from 14 different institutions in academia and the national laboratory complex and is one of three Scientific Discovery through Advanced Computing (SciDAC) institutions currently supported by the Department of Energy's Advanced Scientific Computing Research Program. Two existing SciDAC institutions include:
- FASTMath: Frameworks, algorithms, and scalable technologies for mathematics
- RAPIDS: SciDac Computer Science and Data Institute
Panos Stinis is leader of the Computational Mathematics Group at Pacific Northwest National Laboratory and will become director of the LEADS Institute. “LEADS bridges the gap between scientific machine learning experts and domain scientists, enabling the development of highly customized, accurate, and efficient cutting-edge algorithms that leverage the vast domain knowledge within the Department of Energy complex,” said Stinis.
