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

2025-04-01

AI Summary

This work addresses the challenge of reconstructing the primordial density field from galaxy surveys. We propose a hybrid method integrating standard Baryon Acoustic Oscillation (BAO) reconstruction with deep learning: first recovering large-scale linear features via conventional techniques, then employing a 3D convolutional neural network to learn small-scale nonlinear corrections on subgrids and generalize them across the full survey volume. This constitutes the first seamless coupling of subgrid-scale deep learning with standard reconstruction. The approach enables cross-volume transfer (e.g., from 1 to 3 Gpc³) without retraining and balances accuracy on large scales with robustness in modeling small-scale structure. Evaluated on Quijote N-body simulations under joint configuration- and redshift-space modeling, our method significantly improves the cross-correlation coefficient with the true initial density field and reduces BAO scale measurement errors substantially compared to standard reconstruction. It further exhibits strong robustness against model misspecification and is scalable to next-generation surveys such as DESI.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

We present a hybrid method for reconstructing the primordial density from late-time halos and galaxies. Our approach involves two steps: (1) apply standard Baryon Acoustic Oscillation (BAO) reconstruction to recover the large-scale features in the primordial density field and (2) train a deep learning model to learn small-scale corrections on partitioned subgrids of the full volume. At inference, this correction is then convolved across the full survey volume, enabling scaling to large survey volumes. We train our method on both mock halo catalogs and mock galaxy catalogs in both configuration and redshift space from the Quijote $1(h^{-1},mathrm{Gpc})^3$ simulation suite. When evaluated on held-out simulations, our combined approach significantly improves the reconstruction cross-correlation coefficient with the true initial density field and remains robust to moderate model misspecification. Additionally, we show that models trained on $1(h^{-1},mathrm{Gpc})^3$ can be applied to larger boxes--e.g., $(3h^{-1},mathrm{Gpc})^3$--without retraining. Finally, we perform a Fisher analysis on our method's recovery of the BAO peak, and find that it significantly improves the error on the acoustic scale relative to standard BAO reconstruction. Ultimately, this method robustly captures nonlinearities and bias without sacrificing large-scale accuracy, and its flexibility to handle arbitrarily large volumes without escalating computational requirements makes it especially promising for large-volume surveys like DESI.

Problem

• Reconstruct primordial density from late-time halos and galaxies
• Improve cross-correlation with true initial density field
• Enhance BAO peak recovery accuracy in large surveys

Innovation

• Hybrid BAO and deep learning reconstruction
• Subgrid deep learning correction scaling
• Fisher analysis improves BAO peak recovery