Abstract: The evolution of the universe prior to Big Bang Nucleosynthesis could have gone through a phase of early matter domination which enhanced the growth of small-scale dark matter structure. If this period was long enough, self-gravitating objects formed prior to reheating. We study the evolution of these dense early halos through reheating. At the end of early matter domination, the early halos undergo rapid expansion and eventually eject their matter. We find that this process washes out structure on scales much larger than naively expected from the size of the original halos. We compute the density profiles of the early halo remnants and use them to construct late-time power spectra that include these non-linear effects. We evolve the resulting power spectrum to estimate the properties of microhalos that would form after matter-radiation equality. Surprisingly, cosmologies with a short period of early matter domination lead to an earlier onset of microhalo formation compared to those with a long period. In either case, dark matter structure formation begins much earlier than in the standard cosmology, with most dark matter bound in microhalos in the late universe.

Abstract: The energy distribution of the neutron background was measured for the first time at Hall A of the Canfranc Underground Laboratory. For this purpose we used a novel approach based on the combination of the information obtained with six large high-pressure He-3 proportional counters embedded in individual polyethylene blocks of different size. In this way not only the integral value but also the flux distribution as a function of neutron energy was determined in the range from 1 eV to 10 MeV. This information is of importance because different underground experiments show different neutron background energy dependence. The high sensitivity of the setup allowed to measure a neutron flux level which is about four orders of magnitude smaller that the neutron background at sea level. The integral value obtained is Phi(Hall A) = (3.44 +/- 0.35) x 10(-6) cm(-2) s(-1).

Abstract: The dark matter sector remains completely unknown. It is therefore crucial to keep an open mind regarding its nature and possible interactions. Focusing on the case of Weakly Interacting Massive Particles, in this work we make this general philosophy more concrete by applying modern machine learning techniques to dark matter direct detection. We do this by encoding and decoding the graphical representation of background events in the XENONnT experiment with a convolutional variational autoencoder. We describe a methodology that utilizes the `anomaly score' derived from the reconstruction loss of the convolutional variational autoencoder as well as a pre-trained standard convolutional neural network, in a semi-supervised fashion. Indeed, we observe that optimum results are obtained only when both unsupervised and supervised anomaly scores are considered together. A data set that has a higher proportion of anomaly score is deemed anomalous and deserves further investigation. Contrary to classical analyses, in principle all information about the events is used, preventing unnecessary information loss. Lastly, we demonstrate the reach of learning-focused anomaly detection in this context by comparing results with classical inference, observing that, if tuned properly, these techniques have the potential to outperform likelihood-based methods.