Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

An AI super-resolution field emulator for cosmological hydrodynamics: the Lyman-{\alpha} forest

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read A two-stage deep-learning pipeline emulates high-resolution Lyman-alpha forest baryonic fields from cheap low-resolution simulations, with 1.07% mean power-spectrum error and roughly a 450x speedup.

desk verdict A genuinely useful baryonic extension of the authors' super-resolution framework that delivers on summary-statistic accuracy and speed, but whose subpercent field-level claim is an artifact of dynamic-range normalization. read the letter →

arxiv 2507.16189 v1 pith:RZJA3XP3 submitted 2025-07-22 astro-ph.CO astro-ph.GAastro-ph.IMphysics.comp-ph

classification astro-ph.COastro-ph.GAastro-ph.IMphysics.comp-ph
keywords Lyman-alphaforestsuper-resolutioncosmologicalhydrodynamicsgenerativeadversarialnetworksfieldemulatorintergalacticmediumfluxpowerspectrumdeeplearning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a two-stage deep-learning pipeline can turn a cheap, low-resolution cosmological hydrodynamic simulation into high-resolution baryonic fields—gas density, temperature, velocity, internal energy, and gas/star labels—at redshift $z=3$. The first stage, HydroSR, stochastically super-resolves the low-resolution fields; the second, HydroEmu, deterministically refines them using the simulation's high-resolution initial conditions. Against paired SPH simulations, the authors report sub-percent normalized errors on overdensity, temperature, velocity, and optical depth, a mean relative error of 1.07% in the large-scale Lyman-$\alpha$ flux power spectrum ($k < 3 \times 10^{-2}\,\mathrm{s/km}$), and below-10% error in the flux probability distribution function, all at roughly 450x lower compute cost than the full simulation. If this holds, it would make field-level Lyman-$\alpha$ forest mocks for future surveys feasible at a small fraction of the cost.

What carries the argument

The carrier of the argument is the two-stage generative-adversarial architecture. HydroSR is a hierarchical convolutional generator trained with a Wasserstein GAN with gradient penalty plus supervised Lagrangian and Eulerian losses; it samples stochastic high-resolution eight-channel fields (displacement, velocity, internal energy, and a gas/star label) from the low-resolution input. HydroEmu is a deterministic conditional U-Net—residual blocks with group normalization and SiLU activations—whose 16-channel input concatenates the HydroSR output with eight channels derived from the high-resolution initial conditions, and it is trained with the same composite loss. The U-Net's skip connections preserve fine spatial information, and the high-resolution initial conditions supply the missing small-scale phase information, letting the second stage repair the stochastically generated fields into a realization that matches the paired high-resolution simulation.

What would settle it

Run a genuinely resolving reference simulation from the same initial conditions—for example a $2048^3$ box or a zoom-in with line-of-sight resolution of a few km/s—and compare HydroEmu's flux power spectrum, flux PDF, and field-level maps against it; if the differences at $k \gtrsim 0.1\,\mathrm{s/km}$ or at the pressure-smoothing scale exceed the claimed 1–10% accuracy, the stated errors are reference-limited rather than physical.

Watch

Extended reading notes

Core claim

The paper's central discovery is that conditioning on high-resolution initial conditions lets a deterministic emulator recover the small-scale baryonic structure—down to the roughly 100 kpc pressure-smoothing scale relevant to the Lyman-$\alpha$ forest—that a stochastic super-resolution model alone leaves uncertain. The authors claim that the two-stage model reproduces not only field-level maps but also the temperature–density relation (best-fit $T_0=1.5\times10^4\,\mathrm{K}$, $\gamma=1.41$, against $1.6\times10^4\,\mathrm{K}$ and $1.44$ for the reference) and flux statistics: relative power-spectrum differences below 10% out to $k\sim0.1\,\mathrm{s/km}$, a 1.07% mean relative deviation below $k=3\times10^{-2}\,\mathrm{s/km}$, and flux PDF deviations under 5% over most of the flux range. The implication is that high-resolution baryonic fields and Lyman-$\alpha$ observables can be emulated at field level—not just summarized—at a small fraction of the runtime.

Load-bearing premise

The high-resolution simulation used as ground truth is itself too coarse to fully resolve the Lyman-alpha forest, so the emulator's accuracy is measured against a target that may be smoothed, and matching that target does not by itself guarantee physical fidelity.

Editorial extensions

If this is right

  • The pipeline converts a 287-second low-resolution run into field-level outputs in roughly 594 seconds total, versus 267,000 seconds for the full high-resolution SPH run—a speed-up factor of about 450.
  • Emulated fields support direct synthetic Lyman-alpha spectra and tomographic maps with claimed 0.1–10% accuracy, rather than only summary statistics such as the flux power spectrum.
  • Sub-percent normalized errors on overdensity, temperature, velocity, and optical depth, together with 1.07% mean error on the large-scale flux power spectrum, imply the method is usable for forward-modeling the IGM thermal state and small-scale structure at $z=3$.
  • Because the emulator is conditioned on high-resolution initial conditions, the same framework could in principle be retrained at other redshifts and cosmologies, and the authors identify chunk-wise inference as a route to Gpc-volume mocks for surveys such as DESI.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The reported accuracies measure agreement with a $512^3$ reference that the paper itself calls insufficient to fully resolve the Lyman-alpha forest, so the true small-scale physical fidelity could be lower than the stated sub-percent numbers.
  • Using high-resolution initial conditions as an input ties the method to paired simulations: it cannot super-resolve an arbitrary low-resolution run whose high-resolution ICs are unavailable, which limits direct application to observed or unpaired volumes.
  • The 450x speedup compares GPU inference to CPU SPH; a fairer wall-clock comparison on matched hardware would likely change the factor but probably not the qualitative conclusion.
  • A natural testable extension is to train the two-stage model on multiple cosmologies and redshifts and check whether field-level interpolation degrades gracefully, since the current demonstration is fixed to one cosmology and one snapshot.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper presents a two-stage deep learning framework (HydroSR followed by HydroEmu) that reconstructs high-resolution baryonic fields—density, velocity, internal energy, and gas/star classification—from low-resolution MP-Gadget hydrodynamic simulations conditioned on high-resolution initial conditions, evaluated at z=3 in 50 h^-1 Mpc boxes. The authors report subpercent normalized RMSE for overdensity, temperature, velocity, and optical depth fields; a mean relative error of 1.07% in the large-scale flux power spectrum for k < 3e-2 s/km; less than 10% error in the flux PDF; and a roughly 450x speedup relative to the full SPH simulation. The validation is performed on four held-out simulation volumes against the HR-HydroSim reference, with additional comparisons to observational Ly-alpha forest measurements.

Significance. If the accuracy claims are taken at face value, this would be a practically useful tool for generating Lyman-alpha forest mocks and for field-level emulation at reduced cost. The paper has real strengths: the architecture is sensible, the training/test split over four volumes is a genuine held-out test, the summary-statistic comparisons (P1D, PDF, TDR) are grounded in commonly used diagnostics, and the runtime comparison is concrete. However, the headline field-level claim is not supported by the metric used. The NRMSE normalization in Eq. (5) is dominated by rare extreme pixels, and the paper's own single-sightline NRMSEs and flux decoherence measurements show that small-scale field-level fidelity is far below subpercent. After correcting the presentation, the solid contribution is better stated as an accurate emulator for large-scale flux summary statistics and for visually and statistically plausible field morphology, with clear limitations on pixel-level and small-scale phase fidelity.

major comments (3)
  1. [§2.3, Eq. (5); §3.3; Abstract] The abstract's 'subpercent error for overdensity, temperature, velocity, and optical depth fields' is an artifact of normalizing RMSE by the full dynamic range of the target field. In Eq. (5), NRMSE = RMSE/(x_max - x_min); for these IGM fields the denominator is set by rare extreme pixels (e.g., shock-heated gas up to ~10^7 K, overdensities up to the ~10^3 star-formation threshold). The full-volume RMSE values in §3.3 are 3.89 for overdensity (whose mean is 1), 3.5e4 K for temperature (while the Ly-alpha-absorbing IGM is at ~10^4 K), and 1.3e3 for optical depth; the quoted NRMSEs of 0.34%, 0.28%, 0.59%, and 0.33% therefore do not quantify errors in the pixels that dominate Ly-alpha absorption. The paper's own single-sightline NRMSEs in §3.3 are 1.67%, 6.36%, 1.84%, 3.21%, and 10.0% for the same fields. The authors should either drop the subpercent field-level claim or support it with metrics normalized by the signal (e.g., per-percentile RMSE, standard-deviation normalization, or errors restricted to the density-temperature range relevant to Ly-alpha absorption).
  2. [§2.1; Abstract; §4] The reference used for the small-scale fidelity claim is explicitly acknowledged in §2.1 as not sufficient to fully resolve the Lyman-alpha forest: 'the resolution of the HR-HydroSim runs is not sufficient to fully resolve the Lyman-alpha forest with high accuracy; however, this is acceptable for the purpose of demonstrating the methodology.' Because HydroEmu is a supervised fit to these HR-HydroSim outputs, agreement with held-out HR-HydroSim volumes is a reproduction of the training distribution, not an independent validation of physical fidelity at the pressure-smoothing scale. The abstract's statement that the model 'captures small-scale structures of the intergalactic medium ... down to the 100 kpc pressure smoothing scale' is therefore overclaimed. The text should be reworded to say that the model reproduces the reference HR-HydroSim fields, with the resolution limitation stated as a caveat on physical fidelity.
  3. [§3.5, Fig. 6] The flux decoherence measurement directly contradicts a subpercent field-level fidelity claim. At k = 0.1 s/km—the smallest scale reliably measured in observations cited by the authors—1-r^2(k) is approximately 0.6 for HydroEmu relative to HR-HydroSim, corresponding to a cross-correlation coefficient of only about 0.63. This is an order-unity phase and amplitude mismatch at small scales, not a subpercent error. The good P1D and PDF agreement is not in tension with this because the power spectrum is amplitude-only and the PDF is a one-point statistic, whereas decoherence includes phase information. The paper should present 1-r^2(k) as a primary field-level fidelity metric and temper or remove the field-level subpercent wording in the abstract and Section 4.
minor comments (5)
  1. [§2.1] There is a typo in the paragraph on the quick-Ly-alpha approximation: 'preventing prevent it from dominating' should read 'preventing it from dominating.'
  2. [§2.2, Eq. (3)] The notation in the WGAN-GP objective is inconsistent: the gradient penalty term uses an index 'i' without defining it, and the reader must infer that the gradient is taken with respect to interpolated samples. Please clarify the notation.
  3. [§3.3] The discussion of why single-sightline and full-volume NRMSEs differ attributes the difference only to statistical variation in field values, but the denominator x_max - x_min in Eq. (5) also changes between the two samples; this should be acknowledged explicitly.
  4. [Abstract vs. §4] The abstract claims 'subpercent error' for the fields while the Discussion summarizes the accuracy as '0.1-10% across a range of validation metrics'; these claims should be harmonized after the metric issue is addressed.
  5. [§2.2] The role of the latent noise vector z in the stochastic HydroSR stage is described only in the loss expression; please state explicitly how z is sampled at inference time and whether ensemble predictions from multiple z draws are used in the flux statistics.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the emulator is a supervised surrogate with held-out test volumes and emergent flux observables; rescaling and normalization choices are transparent and not constructional.

full rationale

The paper is an empirical supervised-learning surrogate, not a first-principles derivation, and its central claims are validated out-of-sample. The 20 paired MP-Gadget runs are split 16/4, so the reported RMSE/NRMSE, TDR, P1D, PDF, and decoherence numbers are measured on test volumes not used in training; the model is trained on baryonic particle fields, while the Ly-alpha flux statistics are generated post hoc with the open-source fake-spectra code, so the flux P1D and PDF are emergent rather than fitted targets. The HydroEmu stage is conditioned on HR-HydroICs, but this is an explicit and legitimate input (initial conditions are cheap to generate and are deterministically related to the target only through the simulation code), and the paper does not claim that the LR-HydroSim alone is sufficient. The mean-flux rescaling mentioned in Section 3.5 is a standard calibration and does not force the reported P1D or decoherence values, because those statistics use the mean-normalized contrast deltaF = F/<F> - 1, which is invariant to a constant rescaling of F; it only aligns the first moment of the flux PDF. The NRMSE normalization in Eq. (5) is a transparent, consistently applied metric; although full-volume dynamic ranges make subpercent values easier to achieve than on a single sightline, the paper also reports absolute RMSEs and the same metric on sightlines, so the subpercent claim is a metric statement rather than a hidden fit. Prior-work citations (Ni et al. 2021; Zhang et al. 2025) supply architectures, not load-bearing physical postulates, and no uniqueness theorem is imported. No circular step can be identified.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central empirical claims rest on the fidelity of the training simulations (MP-Gadget quick-Ly-alpha), on the adequacy of the 512^3 reference as ground truth, and on a set of hyperparameters and normalizations (NRMSE dynamic range, mean-flux rescaling) that shape the reported accuracy. No new physical entities are introduced; the model is a data-driven surrogate.

free parameters (5)
  • Neural network weights (HydroSR and HydroEmu) = Not enumerated (millions of weights)
    The model parameters are fitted to minimize the composite loss on 16 training simulation pairs. The central claim of accuracy depends on these fitted weights.
  • Adversarial loss weight lambda_adv = Not stated explicitly (only lambda in WGAN-GP gradient penalty is set to 10)
    Total loss in Eq. (1) combines L2 and adversarial terms with a weighting parameter not reported; the balance affects output fidelity.
  • WGAN-GP gradient penalty coefficient = 10
    Set following common practice (Section 2.2); influences training stability and output distribution.
  • Mean flux rescaling = Chosen to match HR-HydroSim mean flux
    Section 3.5 states both HydroEmu and HR flux fields are rescaled to match the same mean flux before computing power spectrum and PDF; this normalization strongly affects the reported statistics.
  • TDR fit range cuts = -1.0 < log10(rho/rhobar) < 1.0 and 0.1 < log10(T/K) < 5
    Section 3.2 restricts the TDR fit to this regime; the resulting T0 and gamma are diagnostics, not model predictions.
assumptions (4)
  • domain assumption MP-Gadget with the quick-Ly-alpha approximation produces realistic Lyman-alpha forest fields at the resolutions used
    The training and ground-truth data are from this code and approximation (Section 2.1). If this model of the IGM is inaccurate, the emulator inherits the bias.
  • domain assumption At redshift z=99 all particles are in the gas phase, so the HR-IC binary label can be uniformly set to gas
    Used in Stage 2 input construction (Section 2.2). This is physically reasonable but assumes no early star formation or other phase.
  • ad hoc to paper The 512^3 HR-HydroSim is a sufficient reference for the scales claimed (down to 100 kpc pressure smoothing)
    The paper itself notes the HR runs do not fully resolve the forest (Section 2.1), yet the emulator is benchmarked against them.
  • domain assumption Adversarial training with WGAN-GP converges to a generator whose outputs are statistically indistinguishable from real HR fields
    The method relies on GAN convergence for fidelity (Section 2.2); no convergence certificate is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An AI super-resolution field emulator for cosmological hydrodynamics: the Lyman-{\alpha} forest." pith.science (2026). https://pith.science/paper/RZJA3XP3

@misc{pith2026250716189,
  author       = {Pith},
  title        = {Pith review of: An AI super-resolution field emulator for cosmological hydrodynamics: the Lyman-\alpha forest},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RZJA3XP3}},
  note         = {Machine review of arXiv:2507.16189}
}
abstract

We extend our super-resolution and emulation framework for cosmological dark matter simulations to include hydrodynamics. We present a two-stage deep learning model to emulate high-resolution (HR-HydroSim) baryonic fields from low-resolution (LR-HydroSim) simulations at redshift $z = 3$. The method takes as inputs an LR-HydroSim and the high-resolution initial conditions (HR-HydroICs). First, the model stochastically generates high-resolution baryonic fields from the LR-HydroSim. Second, a deterministic emulator refines these fields using HR-HydroICs to reconstruct small-scale structures including displacement, velocity, internal energy, and gas/star classification. Trained on paired low- and high-resolution simulations produced with \texttt{MP-Gadget}, the model captures small-scale structures of the intergalactic medium and %Lyman-$\alpha$ forest observables down to the 100 kpc pressure smoothing scale relevant to the Lyman-$\alpha$ forest. The model achieves subpercent error for overdensity, temperature, velocity, and optical depth fields, a mean relative error of 1.07\% in the large-scale flux power spectrum (\(k < 3 \times 10^{-2}\ \mathrm{s/km}\)), and less than 10\% error in the flux probability distribution function. Notably, the two-stage model reduces the compute time by a factor of $\sim$450 compared to full smoothed particle hydrodynamics at the same resolution. This work demonstrates the potential of this framework as a powerful and efficient tool for generating high-resolution fields offering fast and accurate alternatives to traditional cosmological hydrodynamic simulations and enabling large-volume mock datasets for next-generation cosmological surveys.

Figures

Figures reproduced from arXiv: 2507.16189 by the authors.

Figure 1
Figure 1. Schematic overview of the proposed two-stage framework. The upper panel illustrates the overall model architecture and training process, including Stage 1 (left, purple box) and Stage 2 (right, green box). The lower panels show the detailed generator architectures for each stage: the SR-Generator (yellow box) used in Stage 1 and the Emu-Generator (orange box) used in Stage 2. mates the Wasserstein distance used in t… view at source ↗
Figure 2
Figure 2. Comparison of gas distributions in the LR-HydroSim, HR-HydroSim, and HydroEmu simulations. The left column shows the full simulation volumes, where the color represents temperature and intensity indicates density. A representative sightline is highlighted by the red box and magnified in the right column. The corresponding Lyman-α flux profile along the sightline is overplotted in orange. The HydroEmulator prediction… view at source ↗
Figure 3
Figure 3. Phase-space distribution of gas in the temperature–density plane for the HR-HydroSim (left) and the HydroEmu prediction (right). The dashed black line indicates the best-fit power-law relation to the IGM temperature-density relation, with T0 and γ denoting the temperature at mean density and fit parameter. The color scale shows the logarithmic number of gas elements per bin. We apply this analysis to assess how clos… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Comparison between the HR-HydroSim simulation and the HydroEmu prediction along a randomly selected sightline. The top three panels show the hydrogen overdensity ρ/ρ¯, temperature T, and line-of-sight velocity vy in real (comoving) space, while the bottom two panels sh…
Figure 5
Figure 5. Figure 5: Comparison between HR-HydroSim outputs and HydroEmu predictions for key physical quantities in the Lyman-α forest. Each panel shows a 2D hexbin scatter plot of the HydroEmu prediction versus the HR-HydroSim ground truth for: (a) overdensity ρ/ρ¯, (b) temperature T [K],…
Figure 6
Figure 6. Figure 6: Scale-dependent flux decoherence, 1 − r 2 (k), be￾tween the high-resolution simulation and each of the emula￾tor (HydroEmu, red) and low-resolution simulation (LR-Hy￾droSim, green). Shaded regions show the standard devia￾tion across four test set. The vertical dashed l…
Figure 7
Figure 7. Figure 7: Comparison of the one-dimensional flux power spectra from HydroEmu (solid red) and HR-HydroSim (dashed blue) with observational measurements. Green cir￾cles indicate data from Irˇsiˇc et al. (2017) (XQ-100) at z = 3.0; purple squares represent measurements from Walther…

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Bridging Simulations and EFT: A Hybrid Model of the Lyman-Alpha Forest Field

    astro-ph.CO 2025-12 conditional novelty 6.0 of 10

    A hybrid EFT forward model using N-body displacements reproduces the simulated Lyman-alpha forest to 5% at k <= 1 h/Mpc with a white-noise residual.

Reference graph

Works this paper leans on

58 extracted references · 49 canonical work pages · cited by 1 Pith paper

  1. [1]

    2016, Astronomy & Astrophysics, 594, A27

    Ade, P., Aghanim, N., Arnaud, M., et al. 2016, Astronomy & Astrophysics, 594, A27

  2. [2]

    2016, arXiv preprint arXiv:1611.00036

    Aghamousa, A., Aguilar, J., Ahlen, S., et al. 2016, arXiv preprint arXiv:1611.00036

  3. [3]

    Bagla, J. S. 2002, Journal of Astrophysics and Astronomy, 23, 185

  4. [4]

    2022, Monthly Notices of the Royal Astronomical Society, 000, 1

    Bernardini, M., Feldmann, R., Angl` es-Alc´ azar, D., et al. 2022, Monthly Notices of the Royal Astronomical Society, 000, 1

  5. [5]

    2025, fake–spectra: A toolkit for generating Lyman-alpha spectra from simulations, Available at: https://github.com/sbird/fake spectra

    Bird, S. 2025, fake–spectra: A toolkit for generating Lyman-alpha spectra from simulations, Available at: https://github.com/sbird/fake spectra

  6. [6]

    2022, Monthly Notices of the Royal Astronomical Society, 512, 3703

    Bird, S., Ni, Y., Di Matteo, T., et al. 2022, Monthly Notices of the Royal Astronomical Society, 512, 3703

  7. [7]

    K., Peiris, H

    Bird, S., Rogers, K. K., Peiris, H. V., et al. 2019, Journal of Cosmology and Astroparticle Physics, 2019, 050

  8. [8]

    2023, Journal of Cosmology and Astroparticle Physics, 2023, 037

    Bird, S., Fernandez, M., Ho, M.-F., et al. 2023, Journal of Cosmology and Astroparticle Physics, 2023, 037

Show all 58 references
  1. [9]

    S., & Becker, G

    Bolton, J. S., & Becker, G. D. 2009, Monthly Notices of the Royal Astronomical Society: Letters, 398, L26

  2. [10]

    2023, arXiv preprint arXiv:2303.17939

    Boonkongkird, C., Lavaux, G., Peirani, S., et al. 2023, arXiv preprint arXiv:2303.17939

  3. [11]

    L., Norman, M

    Bryan, G. L., Norman, M. L., O’Shea, B. W., et al. 2014, The Astrophysical Journal Supplement Series, 211, 19

  4. [12]

    2023, Monthly Notices of the Royal Astronomical Society, 525, 3499

    Pedersen, C. 2023, Monthly Notices of the Royal Astronomical Society, 525, 3499

  5. [13]

    2017, The Astrophysical Journal, 837, 19

    Cappelluti, N., Li, Y., Ricarte, A., et al. 2017, The Astrophysical Journal, 837, 19

  6. [14]

    S., Schlegel, D

    Dawson, K. S., Schlegel, D. J., Ahn, C. P., et al. 2012, The Astronomical Journal, 145, 10

  7. [15]

    2019, Monthly Notices of the Royal Astronomical Society, 489, 2536 de Belsunce, R., Ivanov, M

    Day, A., Tytler, D., & Kambalur, B. 2019, Monthly Notices of the Royal Astronomical Society, 489, 2536 de Belsunce, R., Ivanov, M. M., Sullivan, J. M., Akitsu, K., & Chen, S.-F. 2025, arXiv preprint arXiv:2507.00284

  8. [16]

    2018, Neural networks, 107, 3

    Elfwing, S., Uchibe, E., & Doya, K. 2018, Neural networks, 107, 3

  9. [17]

    2024, Journal of Cosmology and Astroparticle Physics, 2024, 029

    Fernandez, M., Bird, S., & Ho, M.-F. 2024, Journal of Cosmology and Astroparticle Physics, 2024, 029

  10. [18]

    Y., & Hui, L

    Gnedin, N. Y., & Hui, L. 1998, Monthly Notices of the Royal Astronomical Society, 296, 44

  11. [19]

    2014, in Proceedings of the 27th International Conference on Neural Information Processing Systems (NeurIPS), 2672–2680

    Goodfellow, I., Pouget-Abadie, J., Mirza, M., et al. 2014, in Proceedings of the 27th International Conference on Neural Information Processing Systems (NeurIPS), 2672–2680

  12. [20]

    Courville, A. C. 2017, Advances in neural information processing systems, 30

  13. [21]

    2012, The Astrophysical Journal, 746, 125

    Haardt, F., & Madau, P. 2012, The Astrophysical Journal, 746, 125

  14. [22]

    2018, The Astronomical Journal, 156, 160

    Hand, N., Feng, Y., Beutler, F., et al. 2018, The Astronomical Journal, 156, 160

  15. [23]

    2022, The Astrophysical Journal, 929, 160

    Harrington, P., Mustafa, M., Dornfest, M., Horowitz, B., & Luki´ c, Z. 2022, The Astrophysical Journal, 929, 160

  16. [24]

    2013, The Astrophysical Journal Supplement Series, 208, 19 15

    Hinshaw, G., Larson, D., Komatsu, E., et al. 2013, The Astrophysical Journal Supplement Series, 208, 19 15

  17. [25]

    Hopkins, P. F. 2015, Monthly Notices of the Royal Astronomical Society, 450, 53

  18. [26]

    Hui, L., & Gnedin, N. Y. 1997, Monthly Notices of the Royal Astronomical Society, 292, 27 Irˇ siˇ c, V., Viel, M., Bolton, T. A., et al. 2017, Physical Review D, 96, 023522

  19. [27]

    Isola, P., Zhu, J.-Y., Zhou, T., & Efros, A. A. 2017, in Proceedings of the IEEE conference on computer vision and pattern recognition, 1125–1134

  20. [28]

    2023, The Astrophysical Journal, 958, 21 Kara¸ caylı, N

    Jacobus, C., Harrington, P., & Luki´ c, Z. 2023, The Astrophysical Journal, 958, 21 Kara¸ caylı, N. G., Martini, P., Guy, J., et al. 2024, Monthly Notices of the Royal Astronomical Society, 528, 3941

  21. [29]

    2007, Monthly Notices of the Royal Astronomical Society, 382, 1657 Kodi Ramanah, D., Charnock, T., Villaescusa-Navarro, F., & Wandelt, B

    Kim, T.-S., Bolton, J., Viel, M., Haehnelt, M., & Carswell, R. 2007, Monthly Notices of the Royal Astronomical Society, 382, 1657 Kodi Ramanah, D., Charnock, T., Villaescusa-Navarro, F., & Wandelt, B. D. 2020, Monthly Notices of the Royal Astronomical Society, 495, 4227

  22. [30]

    2017, in Proceedings of the IEEE conference on computer vision and pattern recognition, 4681–4690

    Ledig, C., Theis, L., Husz´ ar, F., et al. 2017, in Proceedings of the IEEE conference on computer vision and pattern recognition, 4681–4690

  23. [31]

    2011, arXiv preprint arXiv:1104.2932

    Lesgourgues, J. 2011, arXiv preprint arXiv:1104.2932

  24. [32]

    A., et al

    Li, Y., Ni, Y., Croft, R. A., et al. 2021, Proceedings of the National Academy of Sciences, 118, e2022038118

  25. [33]

    List, F., Bhat, I., & Lewis, G. F. 2019, Monthly Notices of the Royal Astronomical Society, 490, 3134 Luki´ c, Z., Stark, C. W., Nugent, P., et al. 2015, Monthly Notices of the Royal Astronomical Society, 446, 3697

  26. [34]

    2016, Annual Review of Astronomy and Astrophysics, 54, 313

    McQuinn, M. 2016, Annual Review of Astronomy and Astrophysics, 54, 313

  27. [35]

    2009, Reviews of Modern Physics, 81, 1405

    Meiksin, A. 2009, Reviews of Modern Physics, 81, 1405

  28. [36]

    2024, Astronomy & Astrophysics, 689, A153

    Nayak, P., Walther, M., Gruen, D., & Adiraju, S. 2024, Astronomy & Astrophysics, 689, A153

  29. [37]

    2021, Monthly Notices of the Royal Astronomical Society, 507, 1021

    Ni, Y., Li, Y., Lachance, P., et al. 2021, Monthly Notices of the Royal Astronomical Society, 507, 1021

  30. [38]

    2022, Monthly Notices of the Royal Astronomical Society, 513, 670 O˜ norbe, J., Garrison-Kimmel, S., Maller, A

    Ni, Y., Di Matteo, T., Bird, S., et al. 2022, Monthly Notices of the Royal Astronomical Society, 513, 670 O˜ norbe, J., Garrison-Kimmel, S., Maller, A. H., et al. 2014, Monthly Notices of the Royal Astronomical Society, 437, 1894

  31. [39]

    2019, Advances in Neural Information Processing Systems, 32

    Paszke, A., Gross, S., Massa, F., et al. 2019, Advances in Neural Information Processing Systems, 32

  32. [40]

    2010, Monthly Notices of the Royal Astronomical Society, 404, 1281

    Katz, N. 2010, Monthly Notices of the Royal Astronomical Society, 404, 1281

  33. [41]

    H., Colombi, S., et al

    Peirani, S., Weinberg, D. H., Colombi, S., et al. 2014, The Astrophysical Journal, 784, 11

  34. [42]

    2022, Monthly Notices of the Royal Astronomical Society, 514, 3222

    Peirani, S., Prunet, S., Colombi, S., et al. 2022, Monthly Notices of the Royal Astronomical Society, 514, 3222

  35. [43]

    2013, Monthly Notices of the Royal Astronomical Society, 428, 540

    Rollinde, E., Theuns, T., Schaye, J., Pˆ aris, I., & Petitjean, P. 2013, Monthly Notices of the Royal Astronomical Society, 428, 540

  36. [44]

    Ronneberger, O., Fischer, P., & Brox, T. 2015, in Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18, Springer, 234–241

  37. [45]

    2002, Annual Review of Astronomy and Astrophysics, 40, 539

    Rosati, P., Borgani, S., & Norman, C. 2002, Annual Review of Astronomy and Astrophysics, 40, 539

  38. [46]

    2024, Astronomy & Astrophysics, 682, A21

    Sinigaglia, F., Kitaura, F.-S., Nagamine, K., Oku, Y., & Balaguera-Antol ´ ınez, A. 2024, Astronomy & Astrophysics, 682, A21

  39. [47]

    Sorini, D., Onorbe, J., Luki´ c, Z., & Hennawi, J. F. 2016, The Astrophysical Journal, 827, 97

  40. [48]

    2005, Monthly Notices of the Royal Astronomical Society, 364, 1105

    Springel, V. 2005, Monthly Notices of the Royal Astronomical Society, 364, 1105

  41. [49]

    2003, Monthly Notices of the Royal Astronomical Society, 339, 289

    Springel, V., & Hernquist, L. 2003, Monthly Notices of the Royal Astronomical Society, 339, 289

  42. [50]

    Sunyaev, R., & Zeldovich, Y. B. 1972, Comments on Astrophysics and Space Physics, Vol. 4, p. 173, 4, 173

  43. [51]

    2002, Astronomy & Astrophysics, 385, 337

    Teyssier, R. 2002, Astronomy & Astrophysics, 385, 337

  44. [52]

    D., Bolton, J

    Viel, M., Becker, G. D., Bolton, J. S., & Haehnelt, M. G. 2013, Physical Review D—Particles, Fields, Gravitation, and Cosmology, 88, 043502

  45. [53]

    G., & Springel, V

    Viel, M., Haehnelt, M. G., & Springel, V. 2004, Monthly Notices of the Royal Astronomical Society, 354, 684

  46. [54]

    2020, Nature Reviews Physics, 2, 42

    Vogelsberger, M., Marinacci, F., Torrey, P., & Puchwein, E. 2020, Nature Reviews Physics, 2, 42

  47. [55]

    F., & Luki´ c, Z

    Walther, M., O˜ norbe, J., Hennawi, J. F., & Luki´ c, Z. 2019, The Astrophysical Journal, 872, 13

  48. [56]

    2018, in Proceedings of the European conference on computer vision (ECCV) workshops, 0–0

    Wang, X., Yu, K., Wu, S., et al. 2018, in Proceedings of the European conference on computer vision (ECCV) workshops, 0–0

  49. [57]

    2025, The Open Journal of Astrophysics, 8

    Zhang, X., Lachance, P., Dasgupta, A., et al. 2025, The Open Journal of Astrophysics, 8

  50. [58]

    2024, Monthly Notices of the Royal Astronomical Society, 528, 281

    Zhang, X., Lachance, P., Ni, Y., et al. 2024, Monthly Notices of the Royal Astronomical Society, 528, 281

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.