Fast Sampling of Cosmological Initial Conditions with Gaussian Neural Posterior Estimation

Abstract

Knowledge of the primordial matter density field from which the large-scale structure of the Universe emerged over cosmic time is of fundamental importance for cosmology. However, reconstructing these cosmological initial conditions from late-time observations is a notoriously difficult task, which requires advanced cosmological simulators and sophisticated statistical methods to explore a multi-million-dimensional parameter space. We show how simulation-based inference (SBI) can be used to tackle this problem and to obtain data-constrained realisations of the primordial dark matter density field in a simulation-efficient way with general non-differentiable simulators. Our method is applicable to full high-resolution dark matter $N$-body simulations and is based on modelling the posterior distribution of the constrained initial conditions to be Gaussian with a diagonal covariance matrix in Fourier space. As a result, we can generate thousands of posterior samples within seconds on a single GPU, orders of magnitude faster than existing methods, paving the way for sequential SBI for cosmological fields. Furthermore, we perform an analytical fit of the estimated dependence of the covariance on the wavenumber, effectively transforming any point-estimator of initial conditions into a fast sampler. We test the validity of our obtained samples by comparing them to the true values with summary statistics and performing a Bayesian consistency test.

arXiv comment9 + 2 pages, 7 figures, 1 table. Comments welcome!

References in corpus (51)

• Array Programming with NumPy
• The cosmological simulation code GADGET-2
• Variational Inference: A Review for Statisticians
• LSST: from Science Drivers to Reference Design and Anticipated Data Products
• The Cosmic Linear Anisotropy Solving System (CLASS) II: Approximation schemes
• Stable clustering, the halo model and nonlinear cosmological power spectra
• Revising the Halofit Model for the Nonlinear Matter Power Spectrum
• The Cosmic Linear Anisotropy Solving System (CLASS) I: Overview
• The Quijote simulations
• Bayesian physical reconstruction of initial conditions from large scale structure surveys
• Memory of Initial Conditions in Gravitational Clustering
• nbodykit: an open-source, massively parallel toolkit for large-scale structure
• Reconstruction of the early Universe as a convex optimization problem
• A reconstruction of the initial conditions of the Universe by optimal mass transportation
• Physical Bayesian modelling of the non-linear matter distribution: new insights into the Nearby Universe
• Velocity correction for Hubble constant measurements from standard sirens
• Iterative initial condition reconstruction
• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• An iterative reconstruction of cosmological initial density fields
• Fast and Credible Likelihood-Free Cosmology with Truncated Marginal Neural Ratio Estimation
• How much information can be extracted from galaxy clustering at the field level?
• Translation and Rotation Equivariant Normalizing Flow (TRENF) for Optimal Cosmological Analysis
• Field Level Neural Network Emulator for Cosmological N-body Simulations
• A neural simulation-based inference approach for characterizing the Galactic Center $\gamma$-ray excess
• Probabilistic Mass Mapping with Neural Score Estimation
• Consistency tests of field level inference with the EFT likelihood
• SICRET: Supernova Ia Cosmology with truncated marginal neural Ratio EsTimation
• Estimating the warm dark matter mass from strong lensing images with truncated marginal neural ratio estimation
• Reconstructing Cosmological Initial Conditions from Late-Time Structure with Convolutional Neural Networks
• On the accuracy and precision of correlation functions and field-level inference in cosmology
• Exploring the posterior surface of the large scale structure reconstruction
• The primordial magnetic field in our cosmic backyard
• The Dark Energy Spectroscopic Instrument (DESI)
• Reconstructing the cosmological density and velocity fields from redshifted galaxy distributions using V-net
• Truncated proposals for scalable and hassle-free simulation-based inference
• Bayesian Inference of Initial Conditions from Non-Linear Cosmic Structures using Field-Level Emulators
• Effective cosmic density field reconstruction with convolutional neural network
• Joint velocity and density reconstruction of the Universe with nonlinear differentiable forward modeling
• Revealing the Local Cosmic Web from Galaxies by Deep Learning
• (DarkAI) Mapping the large-scale density field of dark matter using artificial intelligence
• Improving constraints on primordial non-Gaussianity using neural network based reconstruction
• Reconstructing the Universe with Variational self-Boosted Sampling
• Fast likelihood-free inference in the LSS Stage IV era
• Simple lessons from complex learning: what a neural network model learns about cosmic structure formation
• Diffusion-HMC: Parameter Inference with Diffusion-model-driven Hamiltonian Monte Carlo
• Field-level multiprobe analysis of the CMB, integrated Sachs-Wolfe effect, and the galaxy density maps
• CosmicRIM : Reconstructing Early Universe by Combining Differentiable Simulations with Recurrent Inference Machines
• Learning the Universe: Learning to Optimize Cosmic Initial Conditions with Non-Differentiable Structure Formation Models
• Observational Properties of $\beta$ Cephei Stars: 88 new samples discovered Based on TESS and Gaia Data
• Bayesian imaging inverse problem with scattering transform
• Comparing Mass Mapping Reconstruction Methods with Minkowski Functionals