Trinity Teams Receive DOE Genesis Mission Awards to Advance AI for Scientific Discovery

Two Trinity-led research teams have received Phase I Genesis Mission awards from the U.S. Department of Energy to develop artificial intelligence tools that could accelerate discoveries in nuclear physics and astrophysics, helping scientists better understand matter under extreme conditions and respond more quickly to rare cosmic events. 

The projects, led by Duke physicists Steffen Bass and Kate Scholberg, are among the inaugural awards made through the Department of Energy's Genesis Mission, a national initiative to accelerate scientific discovery by combining artificial intelligence, supercomputing, quantum systems and advanced scientific instruments. The projects bring together collaborators from national laboratories and universities across the country. 

The awardees list also includes two teams led by Duke's Pratt School of Engineering Yiran Chen, the John Cocke Distinguished Professor of Electrical and Computer Engineering, and Gaurav Arya, Professor in the Thomas Lord Department of Mechanical Engineering and Materials Science. 

Bass, Arts & Sciences Distinguished Professor of Physics, will lead a team developing AI tools to better understand the fundamental properties of nuclear matter.  

One of the central challenges in nuclear physics is determining the equation of state of nuclear matter — a universal relation describing how this most basic matter behaves under extreme temperatures and pressures. Understanding this relationship will help scientists explain and predict phenomena ranging from the structure of atomic nuclei to the behavior of stars and supernovae. 

Researchers currently estimate the equation of state by comparing experimental data from heavy-ion collisions or astronomical observations of supernova explosions with sophisticated computer simulations, a process that requires enormous computing resources and relies on researchers selecting which measurements are most likely to reveal the underlying physics. Bass's team will instead develop AI methods that can identify the most informative observables directly from simulation data of heavy-ion collisions, making the process faster, more scalable and less dependent on human intuition. 

"The nuclear equation of state is key to our understanding of many phenomena in nuclear science — from the properties of nuclei across the nuclear chart, to the early universe, to extreme cosmic environments such as neutron stars and supernovae, where many of the elements that make up the world around us are created," Bass said. "We are excited to bring the power of artificial intelligence and machine learning to bear on determining the nuclear equation of state using DOE's flagship accelerator facilities." 

Bass’ collaborators include Duke’s James B. Duke Distinguished Professor of Physics Berndt Mueller, Agnieszka Sorensen at Michigan State University, Ron Soltz and Indra Chakraborty at Lawrence Livermore National Laboratory, and Daniel Cebra at the University of California, Davis. 

Scholberg, Arts & Sciences Distinguished Professor of Physics, will lead a project that uses AI to dramatically improve how quickly scientists can respond to the next nearby supernova. 

The Deep Underground Neutrino Experiment (DUNE) currently under construction nearly a mile underground in South Dakota is designed to study human-made neutrinos produced at Fermilab — the Department of Energy’s particle physics laboratory, located in Illinois. It will also be capable of detecting the burst of neutrinos released when a massive star collapses in a supernova. Neutrinos arrive at Earth before the first visible light from the explosion, so they can provide astronomers with an early warning — potentially minutes or even hours before telescopes would otherwise detect the event. 

The challenge is speed. Extracting the location of the supernova from the enormous volume of detector data is computationally demanding, limiting how quickly astronomers can be alerted. Scholberg's team will develop AI-accelerated methods capable of identifying the direction of an exploding star in real time, allowing observatories around the world to point their telescopes toward the event while it is still in its earliest stages. 

"Neutrino detectors like the Deep Underground Neutrino Experiment provide a unique opportunity to spot an exploding star from deep underground," Scholberg said. "I'm excited to develop ultra-fast AI methods to pinpoint the explosion immediately using the neutrino burst, so the scientific community can make the most of a once-in-a-career fireworks event." 

Scholberg’s collaborators include Tom Junk, Jennifer Ngadiuba and Michael Wang at Fermi National Accelerator Laboratory, and Georgia Karagiorgi at Columbia University. 

The Genesis Mission is a historic national initiative led by the U.S. Department of Energy. By uniting government, industry, academia and philanthropy, it is accelerating breakthroughs in energy, scientific discovery and national security through a new platform that combines AI, supercomputing, quantum systems and advanced scientific instruments. 

The goal of the Phase I awards is to identify promising pathways toward transformative scientific capabilities and establish a foundation for future investment and scale. Project teams will design and demonstrate research workflows that integrate AI with scientific investigation while evaluating whether those approaches can accelerate discovery, improve predictive capabilities, enhance experimentation and generate new scientific insights.  

As Mission awardees, the two Trinity research teams will gain access to the Genesis Mission Platform, including AI agent frameworks, advanced AI models and software made available through industry partners, and high-performance computing resources across DOE’s National Laboratories and partner facilities. Together, these capabilities will enable researchers to rapidly design, test, and refine new approaches to accelerate scientific discovery.