From Atoms to Machine-Learned Molecules: Dark Matter and Neutrino Detection at Low Energies

by Jack Shergold (Durham University)

Thursday 19 Feb 2026, 14:00 → 15:00 Europe/London

 Low mass Dark Matter is an incredibly challenging task, with many models largely unconstrained at MeV masses and below. In order to strongly probe this region, we need detectors with thresholds of eV or smaller, making atoms and molecules ideal candidates. In this talk I will introduce the core principles of using atoms and molecules as DM and neutrino detectors. I will first discuss our work on atoms, introducing the relativistic rate formalism, and CINCO, our program for the fast, automated computation of atomic scattering amplitudes. Next, I will move onto our molecular program, outlining how rates are computed for complex, many-body systems, and how we have fully automated and reduced the computation times to O(s) with SCarFFF. Finally, I will discuss how we plan to navigate the almost infinite space of molecules using machine learning.