Tracing heavy inflationary particles in cosmic structures with deep neural networks

• Jeong Han Kim

Room: 447-1 & ZOOM The production of heavy inflationary particles during inflation has the potential to alter the Large Scale Structure, leading to local enhancements in DM clumps or galaxy densities. We derive the logarithmic shape of the density contrast, predicted by linear perturbation theory, which undergoes significant alterations due to non-linear effects over time. We develop a deep neural network algorithm specifically designed to detect these signals and evaluate the effectiveness of this position space search.

Hubble wars: a nu hope

• Gabriella Barenboim

Room: 447-1 & ZOOM We consider the possibility that the cosmic neutrino background might have a nonthermal spectrum, and investigate its effect on cosmological parameters relative to standard Λ-CDM cosmology. As a specific model, we consider a thermal y-distortion, which alters the distribution function of the neutrino background by depleting the population of low-energy neutrinos and enhancing the high-energy tail. We constrain the thermal y-parameter of the cosmic neutrino background using Cosmic Microwave Background (CMB) and Baryon Acoustic Oscillation (BAO) measurements. The y-parameter increases the number of effective relativistic degrees of freedom, reducing the sound horizon radius and increasing the best-fit value for the Hubble constant H0, substantially reducing the tension between CMB/BAO constraints and direct measurement of the expansion rate from Type-Ia supernovae. Including a spectral distortion also allows for a higher value of the spectral index of scalar fluctuations and relaxes the bound on the sum of the neutrino masses. We note a tension between the Λ-CDM-derived value for the primordial helium abundance and the directly measured value obtained by Pitrou et al from BBN which favors a nonzero neutrino thermal y-parameter. ZOOM Link : https://zoom.us/j/92016034244?pwd=b3FpeFRYSWQrb0QreFR4OTVVNXg4Zz09