Speaker
Description
The systematic discovery of optimal sub-GeV dark matter targets, such as anisotropic molecular crystals, is bottlenecked by the computational cost of evaluating complex many-body scattering form factors. In this talk, I will present a novel machine learning framework designed to overcome this barrier by high-throughput screening through the vastness of material space. I will detail how we couple a generative material discovery model with a bespoke electronic-structure computational framework, to perform fast validation of candidate materials. I will discuss how this pipeline allows us to efficiently identify next-generation detector targets sensitive to the directional dark matter wind and how these calculations have accelerated their deployment in recent experimental efforts.