Likelihood-free in Paris

Towards a Quasi-Universal Field-Level Cosmological Emulator

Confirmed

20 avr. 2022, 16:22

7m

École Normale Supérieure, Paris

Talk (shorter) Talks

Orateur

M. Neerav Kaushal (Michigan Technological University)

Description

We train convolutional neural networks to correct the output of fast and approximate N-body simulations at the field level. Our model, Neural Enhanced COLA, --NECOLA--, takes as input a snapshot generated by the computationally efficient COLA code and corrects the positions of the cold dark matter particles to match the results of full N-body Quijote simulations. We quantify the accuracy of the network using several summary statistics, and find that NECOLA can reproduce the results of the full N-body simulations with sub-percent accuracy down to $k\simeq 1h{\mathrm{M}\mathrm{p}\mathrm{c}}^{-1}$. Furthermore, the model, that was trained on simulations with a fixed value of the cosmological parameters, is also able to extrapolate on simulations with different values of ${\mathrm{\Omega }}_{\mathrm{m}}$, ${\mathrm{\Omega }}_{\mathrm{b}}$, $h$, $n_s$, ${\sigma }_8$, $w$, and $M_{\nu }$ with very high accuracy: the power spectrum and the cross-correlation coefficients are within $\simeq 1\mathrm{\%}$ down to $k=1h{\mathrm{M}\mathrm{p}\mathrm{c}}^{-1}$. Our results indicate that the correction to the power spectrum from fast/approximate simulations or field-level perturbation theory is rather universal. Our model represents a first step towards the development of a fast field-level emulator to sample not only primordial mode amplitudes and phases, but also the parameter space defined by the values of the cosmological parameters.