Equivariant normalizing flows and their application to cosmology
Carolina Cuesta-Lazaro
April 2022 - IAIFI JC
Simulated Data
Data


Prior
Posterior
Forwards
Inverse
Cosmological parameters


EARLY UNIVERSE
LATE UNIVERSE

Normalizing flows: Generative models and density estimators

VAE,GAN ...
Gaussianization


Data space
Latent space

Maximize the data likelihood
NeuralNet
f must be invertible
J efficient to compute
1-D
n-D

Equivariance


Invariance


Equivariant
Invariant
Equivariant
Invariant
Challenge: Expressive + Invertible + Equivariant
1. Continuous time Normalizing flows
ODE solutions are invertible!
z = odeint(self.phi, x, [0, 1])
torchdiffeq
solving the ODE might introduce error in estimating p(x)

Image Credit: https://arxiv.org/abs/1810.01367


Equivariant? GNNs

1. Invertible but expressive
2. Equivariant to E(n)

E(n) equivariant normalizing flows

Cosmological simulations -> Millions of particles!
Solution: Density on mesh + Convolutions in Fourier space
1-D functions learned from data

Cubic splines (8 spline points)
Monotonic rational quadratic splines
(8 spline points)
Loss Function
Generative: Maximize likelihood
Discriminative: target the posterior

Gaussian Random Field:
The Power spectrum is an optimal summary statistic

Analytical likelihood

Flow likelihood

Non-Gaussian N-body simulations
1. Inference



Non-Gaussian N-body simulations
2. Sampling

- Can we quantify the full information content? Can normalizing flows extract all the information there is about cosmology?
- Can the latent space be the initial conditions for the N-body sim?
- Are current models to embed symmetries too constraining?
- Model misspecification?
- Does dimensionality reduction help with interpretability?
deck
By carol cuesta
deck
- 418