Bridging Ab Initio Symmetries and Global Nuclear Masses with Interpretable Neural Networks

arXiv·low signal

Phong Dang, Evander Espinoza, and Xiaoliang Wan use interpretable neural networks to connect ab initio SU(4)/SU(3) nuclear symmetries with global nuclear mass predictions. The interpretability angle is notable, but the domain is nuclear theory with essentially no crossover to AI/developer tooling.