Abstract:A major aim of current and upcoming cosmological surveys is testing deviations from the standard $\Lambda$CDM model, but the full scientific value of these surveys will only be realised through efficient simulation methods that keep up with the increasing volume and precision of observational data. $N$-body simulations of modified gravity (MG) theories are computationally expensive since highly non-linear equations need to be solved to model the non-linear matter evolution; this represents a significant bottleneck in the path to reach the data volume and resolution attained by equivalent $\Lambda$CDM simulations. We develop a field-level, neural-network-based emulator that generates density and velocity divergence fields under the $f(R)$ gravity MG model from the corresponding $\Lambda$CDM simulated fields. Using attention mechanisms and a complementary frequency-based loss function, our model is able to learn this intricate mapping. We further use the idea of latent space extrapolation to generalise our emulator to $f(R)$ models with differing field strengths. The predictions of our emulator agree with the $f(R)$ simulations to within 5% for matter density and to within 10% for velocity divergence power spectra up to $k \sim 2\,h$ $\mathrm{Mpc}^{-1}$. But for a few select cases, higher-order statistics are reproduced with $\lesssim$10% agreement with the $f(R)$ simulations. Latent extrapolation allows our emulator to generalise to different $f(R)$ model variants without explicitly training on those variants. Given a $\Lambda$CDM simulation, the GPU-based emulator is able to reproduce the equivalent $f(R)$ realisation $\sim$600 times faster than full $N$-body simulations. This lays the foundations for a valuable tool for realistic yet rapid mock field generation and robust cosmological analyses.

Multi-fidelity Emulator for Cosmological Large Scale 21 cm Lightcone Images: a Few-shot Transfer Learning Approach with GAN

High-Resolution CMB Lensing Reconstruction with Deep Learning

Emulation of $f(R)$ modified gravity from $Λ$CDM using conditional GANs

Single Remote Sensing Image Super-Resolution Via a Generative Adversarial Network with Stratified Dense Sampling and Chain Training

Two-stage Generative Models of Simulating Training Data at the Voxel Level for Large-Scale Microscopy Bioimage Segmentation

Using conditional GANs for convergence map reconstruction with uncertainties

Super-Resolution Emulation of Large Cosmological Fields with a 3D Conditional Diffusion Model

AI-assisted super-resolution cosmological simulations IV: An emulator for deterministic realizations

An Observationally Driven Multifield Approach for Probing the Circum-Galactic Medium with Convolutional Neural Networks

Cosmological multifield emulator

Simultaneous derivation of galaxy physical properties with multimodal deep learning

Fast emulation of two-point angular statistics for photometric galaxy surveys

Building Cross-Domain Mapping Chains from Multi-CycleGAN for Hyperspectral Image Classification

Superresolution emulation of large cosmological fields with a 3D conditional diffusion model

Deblending Galaxies with Generative Adversarial Networks

Galaxy clustering multi-scale emulation

Deep learning-based radiointerferometric imaging with GAN-aided training

Attention-based Neural Network Emulators for Multi-Probe Data Vectors Part I: Forecasting the Growth-Geometry split

AI-assisted superresolution cosmological simulations

GANSky -- fast curved sky weak lensing simulations using Generative Adversarial Networks

Generative Adversarial Networks recover features in astrophysical images of galaxies beyond the deconvolution limit