Accelerating Nuclear Data Delivery Using AI/ML Approaches in High Complexity Data
The detection of gamma rays and neutrons is key to understanding nuclear reactions and decays that drive stellar lifecycles and energy generation in nuclear reactors. Analysis of the wealth of nuclear science data is often bottlenecked by the need to unfold the nuclear data from the detector response, a computationally- and labor-intensive task.
This project, led by Louisiana State University, seeks to overcome this bottleneck through a two-pronged approach. First, deep neural networks will be developed to enable fast unfolding of detector response functions. Second, an agentic AI will be deployed to assist with the extraction of actual nuclear data from a combination of physical observables from the detector data and previous knowledge as derived from the literature and nuclear property databases.
