florpi
https://florpi.github.io/
IAIFI Fellow
Carol(ina) Cuesta-Lazaro
["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
1-Dimensional
Machine Learning
Secondary anisotropies
Galaxy formation
Intrinsic alignments
DESI, DESI-II, Spec-S5
Euclid / LSST
Simons Observatory
CMB-S4
Ligo
Einstein
xAstrophysics
+ Simulations
[Image Credit: Claire Lamman (CfA/Harvard) / DESI Collaboration]
Base Distribution
Target Distribution
1-to-Many mapping between distributions
Make the data as likely as possible
Prompt
A person half Yoda, half Gandalf
[arXiv:2311.17141]
Base Distribution
Target Distribution
Prompt
Cosmological model
TNG-300
True DM
Inferred DM
Size of training simulation
Galaxy Cluster
Void
1 to Many:
[arXiv:2403.10648]
Prompt
~ Gpc
pc
kpc
Mpc
Gpc
Subgrid Model
[Video credit: Francisco Villaescusa-Navarro]
Gas density
Gas temperature
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
Early Universe Inflation
Late Universe
Energy and matter content
Evolution
Dark matter
Dark energy
Hubble Constant
Baryons
Neutrino masses
Non-Gaussianity
Tilt power spectrum
Multifield Inflation
The Universe's forward model