Speaker
Description
The DMRadio program is developing a suite of experiments to probe QCD axions in the sub-1 μeV mass range, a theoretically motivated regime with implications for grand unified theories and pre-inflationary cosmology. The first of these, DMRadio-50L, is undergoing commissioning and targets axion masses from 5 kHz to 5 MHz. Accessing this low-mass QCD axion regime poses unique experimental challenges, requiring high-field magnets, ultra-sensitive readout, and advanced data analysis. To address these challenges, we implement machine learning–based denoising techniques developed in the ABRACADABRA (ABRA) experiment and integrate them into a flexible analysis framework that leverages statistical inference and signal processing to enhance sensitivity. These tools not only improve axion detection capabilities but also expand the scientific reach of DMRadio-50L. In particular, data from the same detector can be reanalyzed to search for high-frequency gravitational waves and other non-standard signals as demonstrated on the ABRA-10cm. Current results from DMRadio-50L and ABRA-10cm will be presented, with a focus on how machine learning enhances sensitivity across these complementary searches.
