Beyond the Observable Universe

florpi

https://florpi.github.io/

IAIFI Fellow

Carol(ina) Cuesta-Lazaro

\Lambda \mathrm{CDM}
["DESI 2024 VI: Cosmological Constraints from the Measurements of Baryon Acoustic Oscillations" arXiv:2404.03002]

What role did Machine Learning play?

Dark Energy is constant over time

DESI's Dark Energy constraints

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

1-Dimensional

Machine Learning

Secondary anisotropies

Galaxy formation

Intrinsic alignments

DESI, DESI-II, Spec-S5

Euclid / LSST

Simons Observatory

CMB-S4

Ligo

Einstein

The era of Big Data Cosmology

xAstrophysics

 

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

+ Simulations

[Image Credit: Claire Lamman (CfA/Harvard) / DESI Collaboration]

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

Inference a la gradient descent

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

Base Distribution

Target Distribution

1-to-Many mapping between distributions

z \sim p(z)
p(z)
x \sim p(x)
p(x)
\mathcal{L}_\phi = -\sum_i \log p_\phi(x_i)

Make the data as likely as possible

Prompt

A person half Yoda, half Gandalf

[arXiv:2311.17141]

Base Distribution

Target Distribution

Generative models for Galaxy Surveys

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

Prompt

Cosmological model

TNG-300

True DM

Inferred DM

Size of training simulation

Galaxy Cluster

Void

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

p_\phi(\rho_\mathrm{DM}|\rho_\mathrm{Galaxies})

1 to Many:

 [arXiv:2403.10648]

 

Unwinding gravitational evolution

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

Prompt

 ~ Gpc

pc

kpc

Mpc

Gpc

Subgrid Model

How well can we simulate the Universe?

[Video credit: Francisco Villaescusa-Navarro]

Gas density

Gas temperature

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

Subgrid model 1

Subgrid model 2

Subgrid model 3

Subgrid model 4

1. There is a lot of information in large scale structure surveys that ML methods can access

2. We can tackle high dimensional inference problems so far unatainable

3. Theoretical uncertainties are limiting the amount of information we can robustly extract

Data-driven theory

Hybrid simulators, robustness, representation learning

Mapping dark matter, constrained simulations... Let's get creative!

Field level inference

Conclusions

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024

Early Universe Inflation

{\delta_\mathrm{Initial}}

Late Universe

Energy and matter content

Evolution

{\delta_\mathrm{Final}}
\color{darkgray}{\Omega_m}

Dark matter

Dark energy

\color{darkgreen}{w_0, w_a}
\color{darkolive}{H_0}

Hubble Constant

\color{darkred}{\Omega_b}
\color{darkblue}{\sum m_\nu}

Baryons

Neutrino masses

\color{purple}{f_\mathrm{NL}}
\color{darkorange}{n_s}

Non-Gaussianity

Tilt power spectrum

Multifield Inflation

The Universe's forward model

Carolina Cuesta-Lazaro IAIFI/MIT @ Rising Stars in Physics 2024