Emulating Recombination with Neural Networks using Universal Differential Equations

Ben Pennell; James M. Sullivan; Zack Li

REVIEW 3 major objections 6 minor 57 references

Emulating Recombination with Neural Networks using Universal Differential Equations

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that a neural network embedded in an ordinary differential equation solver can emulate the cosmic recombination history with sub-percent accuracy (0.16 percent average difference) across three cosmological parameters…

desk verdict A worthwhile proof-of-concept for UDE-based recombination emulation, but the headline 0.16% accuracy claim cannot be checked until the train/test split and model selection are described. read the letter →

arxiv 2411.15140 v2 pith:Q23QPBIR submitted 2024-11-22 astro-ph.CO

Ben Pennell , Zack Li , James M. Sullivan This is my paper · ORCID

classification astro-ph.CO

keywords cosmicmicrowavebackgroundrecombinationhistoryuniversaldifferentialequationsneuralnetworkemulatorionizationdifferentiableforwardmodelcosmologicalparametersHYREC-2

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

The paper tries to show that a neural network embedded in an ordinary differential equation solver—a Universal Differential Equation—can learn the physics of cosmic recombination directly from data, without the hand-built approximations that go into codes like RECFAST. It trains such a network on 48 ionization histories from HYREC-2, varying the baryon density, matter density, and CMB temperature within ten percent of the Planck 2018 best fit, and reports an average difference of 0.16 percent between the network and the test set. If the claim holds, cosmologists get a fast, differentiable forward model for the ionization history that can be plugged into CMB parameter inference and extended to cosmologies beyond the standard model.

What carries the argument

The Universal Differential Equation: a neural network that takes the state variables $x_H$, $x_{He}$, $T$, the redshift $z$, and the cosmological parameters $\Omega_b$, $\Omega_m$, $T_{CMB}$ as inputs and outputs the derivatives $\dot{x}_H$, $\dot{x}_{He}$, $\dot{T}$. A Runge-Kutta solver of order five integrates these derivatives to produce the ionization history, and the training loss compares this integrated trajectory to the HYREC-2 data, so the network learns the vector field rather than merely interpolating the time series. Training uses batched random subsets of the 48 histories, weight decay, a cosine learning-rate schedule, and selection of the best of 150 random initializations.

What would settle it

Run a fully specified holdout test: train on a fixed subset of the 48 HYREC-2 histories, evaluate on the rest, and report the largest and average percentage difference on the held-out histories as a function of the split; if any held-out history differs by more than about one percent, the sub-percent claim fails. Separately, compare the network's integrated trajectory to HYREC-2 when driven from identical initial conditions at $z=3500$ to confirm the learned vector field, not just the fitted trajectories, is being tested.

Watch

Extended reading notes

Core claim

The central claim is that a Universal Differential Equation—a neural network trained as the right-hand side of an ODE—can reproduce the ionization history of the Universe with sub-percent accuracy over a three-parameter volume around the Planck 2018 cosmology. The network outputs the derivatives of the free hydrogen fraction, free helium fraction, and temperature, and a fifth-order Runge-Kutta solver integrates these from redshift 3500 to 700. Comparing the integrated output to HYREC-2, the paper reports an average difference of 0.16 percent on the test set, which it describes as a comparable emulator of the full recombination physics to the approximations in RECFAST.

Load-bearing premise

The result stands on the claim that the 0.16 percent average difference is measured on histories the network never saw during training and that did not influence which of the 150 random initializations was chosen as final; the paper does not specify how many histories were held out or how the test set was selected.

Editorial extensions

If this is right

Within ten percent of the Planck 2018 best-fit values, the emulator can replace a HYREC-2 call in a CMB analysis and supply gradients of the ionization history with respect to cosmological parameters.
Because the network represents the ODE itself, adding new atomic-physics terms or varying additional parameters is a matter of retraining rather than hand-deriving a new approximate recombination code.
Sub-percent accuracy over the sampled range puts the emulator on par with RECFAST's approximations while covering a wider parameter volume.
The method provides a first step toward automatically constructing recombination emulators for cosmologies beyond $\Lambda$CDM, where the physics is less well known.

Reading between the lines

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

The paper's 0.16 percent figure is an out-of-sample claim, but the test set is never defined; a stricter description of the held-out histories and selection procedure would determine whether the number reflects true generalization.
Since the ionization history depends only on ratios of $\Omega_b$ and $\Omega_m$ to $T_{CMB}^3$, the three-parameter sampling may be redundant; training on the two independent ratios could yield the same accuracy with a smaller input space.
The same UDE construction could be applied to other smooth cosmological forward models, such as the matter power spectrum, where a differentiable ODE-level surrogate would accelerate gradient-based inference.

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.

Desk Editor's Note

read the letter

You should read this one. It is, as far as I can tell, the first application of Universal Differential Equations to recombination history emulation, and it ships public code. The idea is clean: instead of interpolating the ionization history directly, train a neural ODE to reproduce the HYREC-2 derivatives for the three-state system (xH, xHe, T), with redshift and cosmological parameters as inputs. That gives you a differentiable forward model, which is exactly what you want for CMB inference with non-standard cosmologies. Training a neural ODE on stiff recombination dynamics is genuinely nontrivial, and their schedule tricks—cosine learning rate cycles, random batching of 16 out of 48 histories, weight decay sweeps—are the kind of practical detail that makes the paper useful even before the emulator is production-ready. They also deserve credit for the footnote admitting that varying Omega_b, Omega_m, and T_CMB independently is redundant under standard assumptions because the ionization history depends on ratios like Omega_b / T_CMB^3; that kind of honesty is rare.

The soft spots are real but concentrated. The main one is the evaluation protocol. The paper says 48 ionization histories are produced, then calls them "training" histories, then reports an average 0.16% difference against something called a "test set" without ever stating how many histories were held out, how they were chosen, or how the "best" of 150 random initializations was selected. If the test set overlaps the training set, or if the 150-initialization selection used test performance, the sub-percent claim does not demonstrate generalization. Figure 1's caption is also sloppy: the top panel is described as showing the percent difference to "training data" while the bottom panel refers to the "test set," and the relation between the two is unclear. This is not a fatal flaw in the method—it is an under-specified experiment—but it is load-bearing because the entire point of an emulator is to predict unseen cosmologies. A second, milder issue: the paper claims "automatic dimensionality reduction," but the reduction from thousands of atomic levels to three state variables comes from HYREC-2's physics, not from the neural network. That is an overstatement, though they do say "building towards" in the abstract.

Who is this for? Anyone working on fast and differentiable recombination models for CMB-S4 or Simons Observatory-era inference. It is a proof-of-concept, not a product: no timing benchmarks, no downstream likelihood runs, and the parameter range is deliberately narrow. But the recipe is sound and the failure modes of neural ODE training are discussed honestly. If I were editing, I would send it to a referee. The referee should ask for a clear train/test split, a statement of how the best initialization was chosen, and a corrected figure caption. With those fixed, it is a solid methods paper.

Recommendation: engage with it, but insist on the evaluation details before citing the accuracy claim.

Referee Report

3 major / 6 minor

Summary. The paper presents a neural-network ordinary differential equation (UDE) emulator for the cosmic recombination history. Training data are 48 HYREC-2 ionization histories with three varied cosmological parameters (Omega_b, Omega_m, T_CMB) sampled in a Latin hypercube within 10% of the Planck 2018 best-fit values. The network outputs derivatives of the state variables (x_H, x_He, T) as a function of redshift and the cosmological parameters, and a Runge-Kutta solver integrates these derivatives to produce histories. The paper reports an average difference of 0.16% between the network output and a test set, and claims sub-percent accuracy within the sampled parameter range. The authors position this as a first step toward autonomous, differentiable emulators of recombination and provide public code on GitHub.

Significance. If the generalization claim is valid, the paper is a useful proof-of-concept for differentiable surrogate models of recombination that could accelerate CMB parameter inference and be extended beyond LambdaCDM. The public code repository, the use of an external physical code (HYREC-2) for training data, and the explicit acknowledgment of the narrow parameter range are strengths. However, the central sub-percent accuracy claim currently rests on an evaluation protocol that is not fully described, so the significance cannot be fully assessed without clarification.

major comments (3)

[Sections 3.1, 3.2, and 4; Figure 1] The paper's central claim of 0.16% average test-set difference is not backed by a defined test set. Section 3.1 states that 48 ionization histories were used for training and that the network was trained with 150 random initializations, selecting the 'best-performing parameters,' but it does not state how many histories were held out, whether the test set is disjoint from the training set, or what criterion (training loss, validation loss, or test loss) was used to choose the best initialization. Section 3.2 describes random batching of 16 of the 48 histories during training, with no separate validation set. Section 4 reports an average difference against a 'test set' and says it included the Planck 2018 best fits, but because training histories were sampled in a Latin hypercube within 10% of those best-fit values, the Planck best fit could be a training history. If the test histories overlap the training set or if the 150-initialization selection used test-set performance, the 0.16% number does not demonstrate out-of-sample generalization. The authors should specify the split, the selection criterion, and the number of test histories before the sub-percent claim can be evaluated.
[Section 3, Eq. (3.4)] The ODE integration used in training and evaluation is incompletely specified: Eq. (3.4) evaluates the network output as the integral from z_max to z_i, but the paper does not state how the initial condition x(z_max) is obtained. For a forward model intended for new cosmologies, the initial values of x_H, x_He, and T at z=3500 depend on the cosmological parameters (e.g., T_CMB), so a fixed initial condition would bias the prediction. The paper should state whether initial conditions are taken from the HYREC-2 training output, predicted by the network, or supplied externally, and how the same procedure applies to test histories.
[Section 4, Figure 1] The error metric is reported only as a single average percentage difference, with no distribution, number of test points, or definition of the percentage (relative to the HYREC-2 value at each redshift). The figure caption also conflates training and test data: the top panel refers to 'percent difference between the network and training data' while the text describes a test set, further obscuring the evaluation protocol. Please provide the per-history and per-redshift error distribution and clarify what 'average' means.

minor comments (6)

[Figure 1 caption] The caption uses 'training data' and 'training output' in a figure that the text describes as showing test-set performance; please make the terminology consistent.
[Section 3, Eq. (3.1)] The helium mass fraction Y_p is used in Eq. (3.1) but is never defined; please define it explicitly or provide the value used.
[Section 3.4] The normalization and output-scaling constants are not given; please provide the exact transformations or point to the corresponding code to ensure reproducibility.
[Introduction and Section 4] There are several typos, including 'such such as' in the Introduction and 'RECF AST' in the Introduction and Section 4.
[Section 3] The Latin hypercube range is described as 'within ten percent of Planck 2018 best-fit parameters'; please specify whether this is a relative or absolute range for each parameter.
[Reference [53]] The reference for the Adam optimizer lacks bibliographic details; please provide the arXiv identifier or conference information.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the UDE is trained against external HYREC-2 outputs and its reported accuracy is an empirical fit/test comparison, not a quantity defined to equal its inputs.

full rationale

The paper's derivation chain is a standard supervised emulation setup: HYREC-2 supplies 48 ionization histories (Section 3), the UDE is trained by minimizing the L2-regularized L1 loss in Eq. (3.4) between the integrated NN output and those histories, and the reported 0.16% figure in Section 4 is an 'average difference between the NN output and test set.' The emulator output is therefore benchmarked against an external code's output, and no equation or definition makes the prediction equivalent to the training data by construction. The only concern raised by the text is that the test set is never explicitly defined and the choice of the 'best-performing' of 150 random initializations (Section 3.1) is not tied to a described validation split; if test data influenced model selection, the generalization claim would be statistically invalid. That would be a methodological flaw, but the paper provides no direct textual evidence that test data were used in training or selection, so it does not rise to demonstrated circularity under the required standard. There are no load-bearing self-citations: the UDE references [12-15] are external methodological sources, and there is no imported uniqueness theorem or ansatz smuggled in via citation.

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

The central result depends on the trained NN weights and several hand-chosen hyperparameters, plus the assumption that HYREC-2's three-variable output can be modeled as a closed ODE. No new physical entities are introduced.

free parameters (4)

Neural network weights and biases p = not reported (trained values in GitHub repository)
Trained on HYREC-2 ionization histories via Adam; the final trained values are not provided in the paper.
Weight decay coefficient lambda = 0.0001
Selected from {0, 0.0001, 0.001, 0.01, 0.1} by comparing loss curves in Figure 3; a hand-tuned hyperparameter.
Network hyperparameters = 4 hidden layers, 30 neurons per layer, tanh, batch size 16, 150 random initializations
Architecture and training schedule were selected by informal search; no ablation or automated selection is reported.
Input normalization and output scaling constants = not reported
Section 3.4 states inputs are normalized and outputs are scaled, but the exact constants are not given in the paper.

assumptions (4)

domain assumption HYREC-2 outputs are accurate training targets for recombination histories.
The emulator inherits any systematic errors or missing physics in HYREC-2; the paper states the emulator matches HYREC-2 approximations (Section 4).
domain assumption The reduced three-state ODE (xH, xHe, T) with a neural derivative field is a closed dynamical system over z in [700, 3500].
The UDE represents derivatives as a function of only these three states plus parameters; no proof that hidden HYREC-2 substates are unnecessary for closure is given (Section 3, Eq. 3.3).
domain assumption The Latin hypercube sample of 48 histories within 10 percent of Planck 2018 best fits adequately covers the parameter space for generalization.
No convergence or coverage analysis is provided; the paper itself notes only a narrow range is covered (Section 4).
standard math Runge-Kutta 5(4) integration of the learned NN field is sufficiently accurate.
Standard numerical solver; no tolerance or error control details are reported (Section 3).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Emulating Recombination with Neural Networks using Universal Differential Equations." pith.science (2026). https://pith.science/paper/Q23QPBIR

@misc{pith2026241115140,
  author       = {Pith},
  title        = {Pith review of: Emulating Recombination with Neural Networks using Universal Differential Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q23QPBIR}},
  note         = {Machine review of arXiv:2411.15140}
}

abstract

With an aim towards modeling cosmologies beyond the $\Lambda$CDM paradigm, we demonstrate the automatic construction of recombination history emulators while enforcing a prior of causal dynamics. These methods are particularly useful in the current era of precision cosmology, where extremely constraining datasets provide insights into a cosmological model dominated by unknown contents. Cosmic Microwave Background (CMB) data in particular provide a clean glimpse into the interaction of dark matter, baryons, and radiation in the early Universe, but interpretation of this data requires knowledge of the Universe's ionization history. The exploration of new physics with new CMB data will require fast and flexible calculation of this ionization history. We develop a differentiable machine learning model for recombination physics using a neural network ordinary differential equation architecture (Universal Differential Equations, UDEs), building towards automatic dimensionality reduction and the avoidance of manual tuning based on cosmological model.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 15 canonical work pages

[1]

Aghanim, Y

Planck Collaboration, N. Aghanim, Y. Akrami, M. Ashdown, J. Aumont, C. Baccigalupi et al., Planck 2018 results. VI. Cosmological parameters , aap 641 (2020) A6 [ 1807.06209]

arXiv 2020
[2]

Thornton, P.A.R

R.J. Thornton, P.A.R. Ade, S. Aiola, F.E. Angil` e, M. Amiri, J.A. Beall et al., The Atacama Cosmology Telescope: The Polarization-Sensitive ACTPol Instrument , The Astrophysical Journal Supplement Series 227 (2016) 21

work page 2016
[3]

Benson, P.A.R

B.A. Benson, P.A.R. Ade, Z. Ahmed, S.W. Allen, K. Arnold, J.E. Austermann et al., SPT-3G: a next-generation cosmic microwave background polarization experiment on the South Pole telescope, in Millimeter, Submillimeter, and Far-Infrared Detectors and Instrumentation for Astronomy VII, W.S. Holland and J. Zmuidzinas, eds., vol. 9153 of Society of Photo-Opti...

arXiv 2014
[4]

Essinger-Hileman, A

T. Essinger-Hileman, A. Ali, M. Amiri, J.W. Appel, D. Araujo, C.L. Bennett et al., CLASS: the cosmology large angular scale surveyor , in Millimeter, Submillimeter, and Far-Infrared Detectors and Instrumentation for Astronomy VII , W.S. Holland and J. Zmuidzinas, eds., vol. 9153 of Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Serie...

arXiv 2014
[5]

Simons Observatory Collaboration, The Simons Observatory: science goals and forecasts , Journal of Cosmology and Astro-Particle Physics 2019 (2019) 056 [ 1808.07445]

arXiv 2019
[6]

Abazajian, G

K. Abazajian, G. Addison, P. Adshead, Z. Ahmed, S.W. Allen, D. Alonso et al., CMB-S4 Science Case, Reference Design, and Project Plan , arXiv e-prints (2019) arXiv:1907.04473 [1907.04473]

arXiv 2019
[7]

Zeldovich, V.G

Y.B. Zeldovich, V.G. Kurt and R.A. Syunyaev, Recombination of Hydrogen in the Hot Model of the Universe , Zhurnal Eksperimentalnoi i Teoreticheskoi Fiziki 55 (1968) 278

work page 1968
[8]

Peebles, Recombination of the Primeval Plasma , ApJ 153 (1968) 1

P.J.E. Peebles, Recombination of the Primeval Plasma , ApJ 153 (1968) 1

work page 1968

Show all 57 references

[9]

Seager, D.D

S. Seager, D.D. Sasselov and D. Scott, A new calculation of the recombination epoch , The Astrophysical Journal 523 (1999) L1. – 8 –

1999
[10]

Ali-Ha ¨ ımoud and C.M

Y. Ali-Ha ¨ ımoud and C.M. Hirata,HyRec: A fast and highly accurate primordial hydrogen and helium recombination code, Physical Review D 83 (2011)

2011
[11]

Lee and Y

N. Lee and Y. Ali-Ha ¨ ımoud,hyrec-2: A highly accurate sub-millisecond recombination code , Physical Review D 102 (2020)

2020
[12]

Rackauckas, Y

C. Rackauckas, Y. Ma, J. Martensen, C. Warner, K. Zubov, R. Supekar et al., Universal differential equations for scientific machine learning , arXiv preprint arXiv:2001.04385 (2020)

2020 arXiv
[13]

Bolibar, F

J. Bolibar, F. Sapienza, F. Maussion, R. Lguensat, B. Wouters and F. P´ erez, Universal differential equations for glacier ice flow modelling , Geoscientific Model Development 16 (2023) 6671

2023
[14]

Lima, C.M

F.A.R.D. Lima, C.M. Rebello, E.A. Costa, V.V. Santana, M.G.F.d. Moares, A.G. Barreto et al., Improved modeling of crystallization processes by Universal Differential Equations , Chemical Engineering Research and Design 200 (2023) 538

2023
[15]

Vortmeyer-Kley, P

R. Vortmeyer-Kley, P. Nieters and G. Pipa, A trajectory-based loss function to learn missing terms in bifurcating dynamical systems , Scientific Reports 11 (2021) 20394

2021
[16]

Heitmann, D

K. Heitmann, D. Higdon, M. White, S. Habib, B.J. Williams, E. Lawrence et al., The Coyote Universe. II. Cosmological Models and Precision Emulation of the Nonlinear Matter Power Spectrum, ApJ 705 (2009) 156 [ 0902.0429]

2009 arXiv
[17]

T. Auld, M. Bridges and M.P. Hobson, COSMONET: fast cosmological parameter estimation in non-flat models using neural networks , MNRAS 387 (2008) 1575 [ astro-ph/0703445]

2008 arXiv
[18]

Mootoovaloo, A.F

A. Mootoovaloo, A.F. Heavens, A.H. Jaffe and F. Leclercq, Parameter inference for weak lensing using Gaussian Processes and MOPED , MNRAS 497 (2020) 2213 [ 2005.06551]

2020 arXiv
[19]

Spurio Mancini, D

A. Spurio Mancini, D. Piras, J. Alsing, B. Joachimi and M.P. Hobson, COSMOPOWER: emulating cosmological power spectra for accelerated Bayesian inference from next-generation surveys, MNRAS 511 (2022) 1771 [ 2106.03846]

2022 arXiv
[20]

Fendt and B.D

W.A. Fendt and B.D. Wandelt, Pico: Parameters for the Impatient Cosmologist , ApJ 654 (2007) 2 [ astro-ph/0606709]

2007 arXiv
[21]

Knabenhans, J

Euclid Collaboration, M. Knabenhans, J. Stadel, D. Potter, J. Dakin, S. Hannestad et al., Euclid preparation: IX. EuclidEmulator2 - power spectrum emulation with massive neutrinos and self-consistent dark energy perturbations , MNRAS 505 (2021) 2840 [ 2010.11288]

2021 arXiv
[22]

Aric` o, R.E

G. Aric` o, R.E. Angulo and M. Zennaro, Accelerating Large-Scale-Structure data analyses by emulating Boltzmann solvers and Lagrangian Perturbation Theory , arXiv e-prints (2021) arXiv:2104.14568 [2104.14568]

2021 arXiv
[23]

Angulo, M

R.E. Angulo, M. Zennaro, S. Contreras, G. Aric` o, M. Pellejero-Iba˜ nez and J. St¨ ucker,The BACCO simulation project: exploiting the full power of large-scale structure for cosmology , MNRAS 507 (2021) 5869 [ 2004.06245]

2021 arXiv
[24]

Albers, C

J. Albers, C. Fidler, J. Lesgourgues, N. Sch¨ oneberg and J. Torrado, CosmicNet. Part I. Physics-driven implementation of neural networks within Einstein-Boltzmann Solvers , J. Cosmology Astropart. Phys. 2019 (2019) 028 [ 1907.05764]

2019 arXiv
[25]

Jamieson, Y

D. Jamieson, Y. Li, R.A. de Oliveira, F. Villaescusa-Navarro, S. Ho and D.N. Spergel, Field-level Neural Network Emulator for Cosmological N-body Simulations , ApJ 952 (2023) 145 [2206.04594]

2023 arXiv
[26]

Bonici, L

M. Bonici, L. Biggio, C. Carbone and L. Guzzo, Fast emulation of two-point angular statistics for photometric galaxy surveys , arXiv e-prints (2022) arXiv:2206.14208 [ 2206.14208]

2022 arXiv
[27]

Bonici, F

M. Bonici, F. Bianchini and J. Ruiz-Zapatero, Capse.jl: efficient and auto-differentiable CMB power spectra emulation, The Open Journal of Astrophysics 7 (2024) 10 [ 2307.14339]. – 9 –

2024 arXiv
[28]

J. Kwan, S. Saito, A. Leauthaud, K. Heitmann, S. Habib, N. Frontiere et al., Galaxy Clustering in the Mira-Titan Universe. I. Emulators for the Redshift Space Galaxy Correlation Function and Galaxy-Galaxy Lensing , ApJ 952 (2023) 80 [ 2302.12379]

2023 arXiv
[29]

Bolliet, A

B. Bolliet, A. Spurio Mancini, J.C. Hill, M. Madhavacheril, H.T. Jense, E. Calabrese et al., High-accuracy emulators for observables in ΛCDM, Neff , Σmν, and w cosmologies, arXiv e-prints (2023) arXiv:2303.01591 [ 2303.01591]

2023 arXiv
[30]

Moran, K

K.R. Moran, K. Heitmann, E. Lawrence, S. Habib, D. Bingham, A. Upadhye et al., The Mira-Titan Universe - IV. High-precision power spectrum emulation , MNRAS 520 (2023) 3443 [2207.12345]

2023 arXiv
[31]

Boruah, T

S.S. Boruah, T. Eifler, V. Miranda and P.M.S. Krishanth, Accelerating cosmological inference with Gaussian processes and neural networks - an application to LSST Y1 weak lensing and galaxy clustering , MNRAS 518 (2023) 4818 [ 2203.06124]

2023 arXiv
[32]

Eggemeier, B

A. Eggemeier, B. Camacho-Quevedo, A. Pezzotta, M. Crocce, R. Scoccimarro and A.G. S´ anchez,COMET: Clustering observables modelled by emulated perturbation theory , MNRAS 519 (2023) 2962 [ 2208.01070]

2023 arXiv
[33]

G¨ unther, J

S. G¨ unther, J. Lesgourgues, G. Samaras, N. Sch¨ oneberg, F. Stadtmann, C. Fidler et al., CosmicNet II: emulating extended cosmologies with efficient and accurate neural networks , J. Cosmology Astropart. Phys. 2022 (2022) 035 [ 2207.05707]

2022 arXiv
[34]

Kaushal, F

N. Kaushal, F. Villaescusa-Navarro, E. Giusarma, Y. Li, C. Hawry and M. Reyes, NECOLA: Toward a Universal Field-level Cosmological Emulator , ApJ 930 (2022) 115 [ 2111.02441]

2022 arXiv
[35]

Donald-McCann, F

J. Donald-McCann, F. Beutler, K. Koyama and M. Karamanis, MATRYOSHKA: halo model emulator for the galaxy power spectrum , MNRAS 511 (2022) 3768 [ 2109.15236]

2022 arXiv
[36]

Mootoovaloo, A.H

A. Mootoovaloo, A.H. Jaffe, A.F. Heavens and F. Leclercq, Kernel-based emulator for the 3D matter power spectrum from CLASS , Astronomy and Computing 38 (2022) 100508 [2105.02256]

2022 arXiv
[37]

M.-F. Ho, S. Bird and C.R. Shelton, Multifidelity emulation for the matter power spectrum using Gaussian processes, MNRAS 509 (2022) 2551 [ 2105.01081]

2022 arXiv
[38]

Kokron, J

N. Kokron, J. DeRose, S.-F. Chen, M. White and R.H. Wechsler, The cosmology dependence of galaxy clustering and lensing from a hybrid N-body-perturbation theory model , MNRAS 505 (2021) 1422 [ 2101.11014]

2021 arXiv
[39]

A.J. Mead, S. Brieden, T. Tr¨ oster and C. Heymans, HMCODE-2020: improved modelling of non-linear cosmological power spectra with baryonic feedback, MNRAS 502 (2021) 1401 [2009.01858]

2021 arXiv
[40]

Kobayashi, T

Y. Kobayashi, T. Nishimichi, M. Takada, R. Takahashi and K. Osato, Accurate emulator for the redshift-space power spectrum of dark matter halos and its application to galaxy power spectrum, Phys. Rev. D 102 (2020) 063504 [ 2005.06122]

2020 arXiv
[41]

Wibking, D.H

B.D. Wibking, D.H. Weinberg, A.N. Salcedo, H.-Y. Wu, S. Singh, S. Rodr ´ ıguez-Torres et al., Cosmology with galaxy-galaxy lensing on non-perturbative scales: emulation method and application to BOSS LOWZ , MNRAS 492 (2020) 2872 [ 1907.06293]

2020 arXiv
[42]

Nishimichi, M

T. Nishimichi, M. Takada, R. Takahashi, K. Osato, M. Shirasaki, T. Oogi et al., Dark Quest. I. Fast and Accurate Emulation of Halo Clustering Statistics and Its Application to Galaxy Clustering, ApJ 884 (2019) 29 [ 1811.09504]

2019 arXiv
[43]

Zhai, J.L

Z. Zhai, J.L. Tinker, M.R. Becker, J. DeRose, Y.-Y. Mao, T. McClintock et al., The Aemulus Project. III. Emulation of the Galaxy Correlation Function , ApJ 874 (2019) 95 [ 1804.05867]

2019 arXiv
[44]

Bird, K.K

S. Bird, K.K. Rogers, H.V. Peiris, L. Verde, A. Font-Ribera and A. Pontzen, An emulator for the Lyman- α forest, J. Cosmology Astropart. Phys. 2019 (2019) 050 [ 1812.04654]. – 10 –

2019 arXiv
[45]

Schmit and J.R

C.J. Schmit and J.R. Pritchard, Emulation of reionization simulations for Bayesian inference of astrophysics parameters using neural networks , MNRAS 475 (2018) 1213 [ 1708.00011]

2018 arXiv
[46]

Lawrence, K

E. Lawrence, K. Heitmann, J. Kwan, A. Upadhye, D. Bingham, S. Habib et al., The Mira-Titan Universe. II. Matter Power Spectrum Emulation , ApJ 847 (2017) 50 [ 1705.03388]

2017 arXiv
[47]

J. Kwan, K. Heitmann, S. Habib, N. Padmanabhan, E. Lawrence, H. Finkel et al., Cosmic Emulation: Fast Predictions for the Galaxy Power Spectrum , ApJ 810 (2015) 35 [ 1311.6444]

2015 arXiv
[48]

Agarwal, F.B

S. Agarwal, F.B. Abdalla, H.A. Feldman, O. Lahav and S.A. Thomas, PkANN - II. A non-linear matter power spectrum interpolator developed using artificial neural networks , MNRAS 439 (2014) 2102 [ 1312.2101]

2014 arXiv
[49]

Heitmann, E

K. Heitmann, E. Lawrence, J. Kwan, S. Habib and D. Higdon, The Coyote Universe Extended: Precision Emulation of the Matter Power Spectrum , ApJ 780 (2014) 111 [ 1304.7849]

2014 arXiv
[50]

Lawrence, K

E. Lawrence, K. Heitmann, M. White, D. Higdon, C. Wagner, S. Habib et al., The Coyote Universe. III. Simulation Suite and Precision Emulator for the Nonlinear Matter Power Spectrum, ApJ 713 (2010) 1322 [ 0912.4490]

2010 arXiv
[51]

Ivanov, Y

M.M. Ivanov, Y. Ali-Ha ¨ ımoud and J. Lesgourgues,H0 tension or T 0 tension?, Phys. Rev. D 102 (2020) 063515 [ 2005.10656]

2020 arXiv
[52]

Tsitouras, Runge–kutta pairs of order 5(4) satisfying only the first column simplifying assumption, Computers & Mathematics with Applications 62 (2011) 770

C. Tsitouras, Runge–kutta pairs of order 5(4) satisfying only the first column simplifying assumption, Computers & Mathematics with Applications 62 (2011) 770

2011
[53]

Kingma and J

D.P. Kingma and J. Ba, Adam: A method for stochastic optimization , 2017

2017
[54]

Rackauckas, M

C. Rackauckas, M. Innes, Y. Ma, J. Bettencourt, L. White and V. Dixit, Diffeqflux.jl - A julia library for neural differential equations , CoRR abs/1902.02376 (2019) [ 1902.02376]

2019 arXiv
[55]

Mert Turan and J

E. Mert Turan and J. J¨ aschke,Multiple shooting for training neural differential equations on time series , arXiv e-prints (2021) arXiv:2109.06786 [ 2109.06786]

2021 arXiv
[56]

Seager, D.D

S. Seager, D.D. Sasselov and D. Scott, How exactly did the universe become neutral? , The Astrophysical Journal Supplement Series 128 (2000) 407

2000
[57]

Hazumi, P.A.R

M. Hazumi, P.A.R. Ade, A. Adler, E. Allys, K. Arnold, D. Auguste et al., LiteBIRD satellite: JAXA’s new strategic L-class mission for all-sky surveys of cosmic microwave background polarization, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series , v...

2020

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Aghanim, Y

Planck Collaboration, N. Aghanim, Y. Akrami, M. Ashdown, J. Aumont, C. Baccigalupi et al., Planck 2018 results. VI. Cosmological parameters , aap 641 (2020) A6 [ 1807.06209]

arXiv 2020

[2] [2]

Thornton, P.A.R

R.J. Thornton, P.A.R. Ade, S. Aiola, F.E. Angil` e, M. Amiri, J.A. Beall et al., The Atacama Cosmology Telescope: The Polarization-Sensitive ACTPol Instrument , The Astrophysical Journal Supplement Series 227 (2016) 21

work page 2016

[3] [3]

Benson, P.A.R

B.A. Benson, P.A.R. Ade, Z. Ahmed, S.W. Allen, K. Arnold, J.E. Austermann et al., SPT-3G: a next-generation cosmic microwave background polarization experiment on the South Pole telescope, in Millimeter, Submillimeter, and Far-Infrared Detectors and Instrumentation for Astronomy VII, W.S. Holland and J. Zmuidzinas, eds., vol. 9153 of Society of Photo-Opti...

arXiv 2014

[4] [4]

Essinger-Hileman, A

T. Essinger-Hileman, A. Ali, M. Amiri, J.W. Appel, D. Araujo, C.L. Bennett et al., CLASS: the cosmology large angular scale surveyor , in Millimeter, Submillimeter, and Far-Infrared Detectors and Instrumentation for Astronomy VII , W.S. Holland and J. Zmuidzinas, eds., vol. 9153 of Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Serie...

arXiv 2014

[5] [5]

Simons Observatory Collaboration, The Simons Observatory: science goals and forecasts , Journal of Cosmology and Astro-Particle Physics 2019 (2019) 056 [ 1808.07445]

arXiv 2019

[6] [6]

Abazajian, G

K. Abazajian, G. Addison, P. Adshead, Z. Ahmed, S.W. Allen, D. Alonso et al., CMB-S4 Science Case, Reference Design, and Project Plan , arXiv e-prints (2019) arXiv:1907.04473 [1907.04473]

arXiv 2019

[7] [7]

Zeldovich, V.G

Y.B. Zeldovich, V.G. Kurt and R.A. Syunyaev, Recombination of Hydrogen in the Hot Model of the Universe , Zhurnal Eksperimentalnoi i Teoreticheskoi Fiziki 55 (1968) 278

work page 1968

[8] [8]

Peebles, Recombination of the Primeval Plasma , ApJ 153 (1968) 1

P.J.E. Peebles, Recombination of the Primeval Plasma , ApJ 153 (1968) 1

work page 1968