Hybrid Physical-Neural Simulator for Fast Cosmological Hydrodynamics

TL;DR

This paper introduces a hybrid physical-neural simulator for cosmological hydrodynamics that combines differentiable gravity solvers with neural network parametrized hydrodynamics, enabling efficient inference from limited data.

Contribution

A novel hybrid approach integrating differentiable particle-mesh gravity with neural network hydrodynamics for fast, data-efficient cosmological simulations.

Findings

Outperforms existing methods like Enthalpy Gradient Descent.

Requires only a single simulation to train the neural pressure model.

Potential for direct fitting to observational data.

Abstract

Cosmological field-level inference requires differentiable forward models that solve the challenging dynamics of gas and dark matter under hydrodynamics and gravity. We propose a hybrid approach where gravitational forces are computed using a differentiable particle-mesh solver, while the hydrodynamics are parametrized by a neural network that maps local quantities to an effective pressure field. We demonstrate that our method improves upon alternative approaches, such as an Enthalpy Gradient Descent baseline, both at the field and summary-statistic level. The approach is furthermore highly data efficient, with a single reference simulation of cosmological structure formation being sufficient to constrain the neural pressure model. This opens the door for future applications where the model is fit directly to observational data, rather than a training set of simulations.