<ul><li>Researchers have presented a hybrid method for reconstructing the primordial density from late-time halos and galaxies.</li><li>The method involves applying standard Baryon Acoustic Oscillation (BAO) reconstruction to recover large-scale features in the primordial density field.</li><li>A deep learning model is trained to learn small-scale corrections on partitioned subgrids of the full volume.</li><li>The approach significantly improves the reconstruction cross-correlation coefficient with the true initial density field and enables scaling to large survey volumes.</li></ul>

Initial Conditions from Galaxies: Machine-Learning Subgrid Correction to Standard Reconstruction

Discover more