TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

research

One Hundred Simulations Reached Beyond Flat Cosmology

A two-stage halo emulator used evolution mapping to extend percent-level predictions into dynamical dark-energy tests.

Published Updated Story ID: mp-2026-09-13-019
Read the complete editionStory JSON

Summary

A two-stage halo emulator used evolution mapping to extend percent-level predictions into dynamical dark-energy tests.

Aletheia separates parameters that shape the linear power spectrum from those that rescale its evolution, then predicts the halo mass function with two Gaussian-process stages. It trained on 100 flat-ΛCDM simulations and was tested both on independent runs and a dynamical-dark-energy suite. The authors report percent-level accuracy across a wide mass and redshift range, with modest degradation at early times. The public emulator extends the tested framework; accuracy outside its validation domain still requires checking.

Why it matters

A two-stage halo emulator used evolution mapping to extend percent-level predictions into dynamical dark-energy tests.

Limits and context

No additional limitation was separately recorded.

Key claims

  1. A two-stage halo emulator used evolution mapping to extend percent-level predictions into dynamical dark-energy tests.

    Evidence: source-2026-09-13-021

Sources

  1. arXiv preprint 2609.10665arXiv · primary research

Corrections

No corrections have been recorded for this story.