Normalizing flows for random fields in cosmology
arXiv:2105.12024
Abstract
Normalizing flows are a powerful tool to create flexible probability distributions with a wide range of potential applications in cosmology. Here we are studying normalizing flows which represent cosmological observables at field level, rather than at the level of summary statistics such as the power spectrum. We evaluate the performance of different normalizing flows for both density estimation and sampling of near-Gaussian random fields, and check the quality of samples with different statistics such as power spectrum and bispectrum estimators. We explore aspects of these flows that are specific to cosmology, such as flowing from a physical prior distribution and evaluating the density estimation results in the analytically tractable correlated Gaussian case.
References in corpus (17)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- Variational Inference with Normalizing Flows
- Weak Gravitational Lensing of the CMB
- Acoustic Signatures in the Primary Microwave Background Bispectrum
- The frontier of simulation-based inference
- Density estimation using Real NVP
- Normalizing Flows for Probabilistic Modeling and Inference
- Fast likelihood-free cosmology with neural density estimators and active learning
- Masked Autoregressive Flow for Density Estimation
- Review of local non-Gaussianity from multi-field inflation
- Optimal Estimation of Non-Gaussianity
- Maximum a posteriori CMB lensing reconstruction
- Reconstruction of the remote dipole and quadrupole fields from the kinetic Sunyaev Zel'dovich and polarized Sunyaev Zel'dovich effects
- Rapid Separable Analysis of Higher Order Correlators in Large Scale Structure
- KSZ tomography and the bispectrum
- A Generative Model of Galactic Dust Emission Using Variational Inference
- Flow-Based Likelihoods for Non-Gaussian Inference