The new AI-powered approach, Simulation-Based Inference of Galaxies (SimBIG), has significantly improved the precision of cosmological parameters, halving the uncertainty of some values.
ChangHoon Hahn notes, 'We haven't been able to go down to small scales. For a couple of years now, we've known there's additional information there; we just didn't have a good way of extracting it.'
The AI model effectively learned from 2,000 simulated universes, revealing how different cosmological parameters influence galaxy clustering, making analysis more efficient and precise.
The ability to analyze real galaxy observations from the Baryon Oscillation Spectroscopic Survey with such precision signifies a major advancement in astrophysics through the use of AI.
Collection
[
|
...
]