Ab initio emulation of atomic nuclei: Bayesian inference for unraveling the complexities of the strong nuclear interaction

Research Project, 2025 – 2028

The goal of this high-impact project is realistic ab initio predictions of the binding, low-energy quantum structure, and electromagnetic transitions in key exotic nuclei to learn the limits of nuclear stability. We will primarily target light-mass nuclei and medium-mass Neon, Oxygen, Magnesium, Calcium, Nickel, and Tin isotopes that are pivotal in forefront explorations at new radioactive ion beam facilities worldwide.The primary objective is therefore to develop a realistic and novel fourth-order nuclear interaction in chiral effective field theory via pioneering emulator-driven Bayesian methods and new ab initio many-body theory.Expected outcomes:a fundamental description of the strong nuclear interaction grounded in QCD and with unprecedented accuracy and quantified uncertaintyrealistic ab initio predictions for exotic nuclei crucial for delineating dripline physicsab initio emulators to leverage machine learning, model mixing, and density functional theories of nucleiinsights into the stability of visible matterThe breakthroughs that permit this project to succeed are:eigenvector continuation emulators that accurately mimic ab initio calculations millions of times faster than exact ab initio simulationsrecent ab initio methods now capable of capturing multiscale physics, i.e., bulk properties, deformed nuclear states, and electromagnetic momentspowerful Bayesian techniques and sampling strategies for quantifying theoretical uncertainties