EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Benchmarking AI-evolved cosmological structure formation

Xiaofeng Dong, Nesar Ramachandra, Salman Habib, Katrin Heitmann

Monthly Notices of the Royal Astronomical Society · 14.11.2026

DOI: 10.1093/mnras/stag2032 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Monthly Notices of the Royal Astronomical Society, online erschienen am 14.11.2026. Verfasst von Xiaofeng Dong und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The potential of deep learning-based image-to-image translations has attracted significant attention as a fast, approximate alternative to cosmological simulations, which would be particularly useful in various contexts, including covariance studies, investigations of systematics, and cosmological parameter inference. We model the evolution of structure formation using U-Net, a widely employed convolutional image translation framework with previous applications in cosmology. To study aspects of learning-based cosmological mappings in a controlled, illustrative setting, we use two approaches for generating suitable cosmological matter fields as datasets: a simple analytical prescription provided by the Zel’dovich approximation, and a standard N-body method using Particle-Mesh simulations. We apply established metrics such as matter power spectra and bispectra, topological indicators, and statistical independence of density fields and study their reproducibility and convergence properties. The U-Net approach performs well only for some of these physical metrics, and accuracy is worse at more nonlinear scales, where the dynamic range in density is larger. Motivated by field-level emulator studies utilizing Eulerian-aware loss functions and beyond pixel-wise mean squared errors, we examine a density-weighted objective. In the PM tests ending at z = 4, where the evolution is only mildly nonlinear, using the Eulerian matter density evaluated at each particle position as a direct weight on that particle’s Lagrangian displacement error improves the U-Net results at smaller scales; this result does not establish comparable accuracy in the strongly nonlinear low-redshift regime.

Zitieren

APA

Dong, X., Ramachandra, N., Habib, S., & Heitmann, K. (2026). Benchmarking AI-evolved cosmological structure formation. Monthly Notices of the Royal Astronomical Society. https://doi.org/10.1093/mnras/stag2032

Vancouver

Dong X, Ramachandra N, Habib S, Heitmann K. Benchmarking AI-evolved cosmological structure formation. Monthly Notices of the Royal Astronomical Society. 2026. doi:10.1093/mnras/stag2032

BibTeX
@article{Dong2026,
  author = {Dong, Xiaofeng and Ramachandra, Nesar and Habib, Salman and Heitmann, Katrin},
  title = {Benchmarking AI-evolved cosmological structure formation},
  journal = {Monthly Notices of the Royal Astronomical Society},
  year = {2026},
  doi = {10.1093/mnras/stag2032}
}
Kostenlos starten →