PatchNet: A hierarchical approach for neural field-level inference from Quijote Simulations

Anirban Bairagi; Benjamin Wandelt

Home/Archive/arXiv

arXiv · 2509.03165

PatchNet: A hierarchical approach for neural field-level inference from Quijote Simulations

Anirban Bairagi·Benjamin Wandelt

View original source ↗Read source document ↗Export metadata ↓

Abstract

\textit{What is the cosmological information content of a cubic Gigaparsec of dark matter? } Extracting cosmological information from the non-linear matter distribution has high potential to tighten parameter constraints in the era of next-generation surveys such as Euclid, DESI, and the Vera Rubin Observatory. Traditional approaches relying on summary statistics like the power spectrum and bispectrum, though analytically tractable, fail to capture the full non-Gaussian and non-linear structure of the density field. Simulation-Based Inference (SBI) provides a powerful alternative by learning directly from forward-modeled simulations. In this work, we apply SBI to the \textit{Quijote} dark matter simulations and introduce a hierarchical method that integrates small-scale information from field sub-volumes or \textit{patches} with large-scale statistics such as power spectrum and bispectrum. This hybrid strategy is efficient both computationally and in terms of the amount of training data required. It overcomes the memory limitations associated with full-field training. We show that our approach enhances Fisher information relative to analytical summaries and matches that of a very different approach (wavelet-based statistics), providing evidence that we are estimating the full information content of the dark matter density field at the resolution of $\sim 7.8~\mathrm{Mpc}/h$.

Explore related subjects

astro-ph.CO cs.IT math.IT

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Anirban Bairagi, Benjamin Wandelt. 2025-09-03. PatchNet: A hierarchical approach for neural field-level inference from Quijote Simulations. https://arxiv.org/abs/2509.03165

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Primordial black hole clustering from spectator fields for interpreting the JWST observations

The observations by the James Webb Space Telescope (JWST) have revealed unexpectedly massive galaxy candidates at high redshifts, posing a significant challenge to the $Λ$CDM model. In this work, we investigate whether primordial black holes (PBHs) with spatial clustering, generated by a light spectator field during inflation, can accelerate early structure formation. We adopt the galaxy candidates with inferred stellar mass $10^9\,M_\odot\leq M_*^{\rm obs}\leq10^{11}\,M_\odot$ at redshift $7 \leq z \leq 10$ reported by the CEERS program as a benchmark. Two different mechanisms are considered, through which PBH clustering can influence structure formation: the PBH-induced isocurvature perturbations that enhance the matter power spectrum on linear scales, and the localized seed formation and accretion by compact PBH clusters on nonlinear scales. We find that, when adopting the cosmic microwave background (CMB) isocurvature constraint $β_{\rm iso}<0.035$ at the benchmark pivot scale $k_*=0.002\,{\rm Mpc}^{-1}$, PBH clustering can produce a cumulative stellar mass density consistent with the JWST observations while satisfying the relevant isocurvature constraint. However, the allowed enhancement of structure formation is strongly suppressed when the constraint at $k_*=0.1\,{\rm Mpc}^{-1}$ is imposed, indicating a significant dependence on the choice of the pivot scale. In contrast, the localized seed effect of compact PBH clusters is strongly constrained by the CMB isocurvature bounds, while isolated supermassive PBHs produce stellar mass densities far below those inferred from the JWST observations. Our results show that PBH clustering induced by a spectator field can substantially accelerate early structure formation, but whether it can fully account for the JWST-inferred stellar mass density depends sensitively on the pivot scale adopted for the CMB isocurvature constraint.

astro-ph.CO↗

arXiv2026

Probing memory-burdened Primordial Black Holes with global 21 cm signal

We investigate the imprints of memory-burdened primordial black holes (PBH) on the global 21 cm signal during the cosmic dawn. Recent studies reopened the possibility of a mass window of PBHs as a compelling candidate for dark matter, particularly in low-mass regimes ($M_{\text {PBH}}< 10^{15}$ g) where conventional constraints from evaporation are being revisited in light of quantum gravitational effects. One such effect, the \textit{memory burden effect}, slows down black hole evaporation by incorporating the backreaction of radiation on the black hole microstates, substantially extending the lifetime of light PBHs and thus modifying their late-time emission spectra. This prolonged emission can dramatically alter the energy injection history in the early universe. By computing the modified energy injection rates into the intergalactic medium and incorporating them into the thermal and ionization evolution of neutral hydrogen, we obtain projected constraints on the fraction of dark matter. The bounds are obtained from the fact that these low mass PBHs, which were thought otherwise evaporated, can modify the absorption amplitude in the global 21-cm signal at redshift $z\approx17$. Considering the two viable scenarios of transition to the memory-burden phase: fast (or instantaneous) and slow (transition with a finite width), we show how the 21 cm bounds are sensitive to different mass ranges. For a broad transition with $δ=10^{-2}$ we find that PBHs in the mass range $M_{\rm PBH}\simeq10^{8}$-$10^{13}$g are excluded at the level of $f_{\rm PBH}\gtrsim10^{-8}$. In contrast, for a fast-transition case with the lowest suppression exponent $k=1$, the evaporation is suppressed so efficiently that no meaningful 21\,cm constraint remains for $M_{\rm PBH}\gtrsim10^{7}$g.

astro-ph.CO↗

arXiv2026

Axion Inflation with a Massive Abelian Gauge Field

An axial coupling between an inflaton and an Abelian gauge field can trigger the tachyonic amplification of one gauge-field helicity. For a massless vector, modes with physical momentum $k/a\sim |ξ|H$ are enhanced by approximately $\exp(π|ξ|)$, and sufficiently efficient production can provide substantial friction for the homogeneous inflaton. We extend this mechanism to a vector of mass $m$. The instability is present only for $|ξ|>\bar m\equiv m/H$, and in the heavy regime the mode amplitude scales as $\exp[π(|ξ|-\bar m)]$. Because the amplified modes remain well inside the Hubble radius when $\bar m\gg1$, their contribution to long-wavelength curvature perturbations $ζ$ is power-law suppressed at fixed background backreaction. In the weak-backreaction regime we obtain a spectrum ${\cal P}^{\rm id}_ζ \propto \bar m^{-w}$, with $w \sim 2$, while including the gauge-induced friction of scalar perturbations gives the scaling ${\cal P}^{\rm id}_ζ\propto \bar{m}^{-s}$, with $s$ between three and four. These estimates indicate that ${\cal P}^{\rm id}_ζ\lesssim 10^{-9}$ on CMB scales should be compatible with gauge field backreaction for $\bar{m}$ larger than order a few hundred. We test the analytical mode functions and backreaction estimates with the first lattice simulations based on a massive-vector extension of the \texttt{Pencil Code}, including simulations in the strongly backreacting regime.

astro-ph.CO↗