Oral Presentation

Field level inference: from weak lensing to galaxies

Author(s): Uros Seljak (UC Berkeley)

Presenter: Uros Seljak (UC Berkeley)

Cosmological field level inference (FLI) aims to extract maximal information from survey data using information encoded in the two point correlations and beyond. In recent years there has been a lot of development of AI powered solutions to FLI. Simulation Based Inference uses simulations and Neural Networks to learn the likelihoods, or the mapping between data and initial conditions, and this approach has been shown to extract more information than traditional two point correlation analysis. I will present examples to both weak lensing and galaxy clustering.