"Uncovering Dark Matter with Compact Objects and Automatic Differentiation"
The nature of dark matter (DM) in the Universe remains one of the great open questions of particle astrophysics and cosmology today. The WIMP (weakly interacting massive particle) DM paradigm has fallen, leaving us with a wide range of possible DM models and signatures. New methods and ideas are required to efficiently progress. I will discuss ongoing searches for axion DM signatures using radio observations of neutron stars, and discuss the potential role of black holes and gravitational waves. I will furthermore demonstrate how machine learning technology like automatic differentiation and deep universal probabilistic programming can significantly improve the analysis of astrophysical data. As application I will present preliminary results for the analysis of strongly lensed images.