A flexible Bayesian framework for atomic masses by locally inferring configuration mixing

Scientists have developed a new probabilistic model called SPICE to more accurately predict atomic masses and their uncertainties. This framework utilizes Bayesian averaging to account for the mixing of different low-lying nuclear configurations. By considering excitations across major shells, the model achieves high accuracy with a limited number of parameters. SPICE demonstrates improved predictive power and offers insights into how nuclear shell structures change in proton- and neutron-rich environments.

Accurate atomic mass predictions are crucial for understanding the creation of heavy elements in cosmic events and for advancing nuclear physics research.