AI and the World's Hard Problems · Physics · arXiv · 2026-10-26
Machine Learning $\beta$-decay Half-lives and Their Application to $r$-Process Observables
Accurately modelling β-decay half-lives of neutron-rich nuclei is important for r-process nucleosynthesis and kilonova light curves. A Mixture Density Network trained on recent experimental data outputs a probability distribution of half-lives, while varying the training data produces a wide spread of extrapolations, especially in data-poor regions such as the N=126 isotonic chain. The resulting models were also propagated into r-process network calculations to assess their effect on isotopic abundances and light curves.