pith. sign in

arxiv: 2606.10024 · v1 · pith:HJ47I5TRnew · submitted 2026-06-08 · 🌌 astro-ph.CO · astro-ph.GA

Learning the Universe with cosmological rescaling of merger trees and semi-analytic galaxy formation models

Richard Stiskalek , Lucia A. Perez , Shy Genel , Rachel S. Somerville , Raul E. Angulo , Sergio Contreras This is my paper

Pith reviewed 2026-06-27 15:18 UTC · model grok-4.3

classification 🌌 astro-ph.CO astro-ph.GA
keywords cosmological rescalingmerger treessemi-analytic modelsgalaxy formationN-body simulationscosmological parametersOmega_msigma_8
0
0 comments X

The pith

Rescaling merger trees from 64 base N-body simulations produces galaxy statistics across cosmological parameters at the accuracy of 750 full simulations.

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

The paper establishes that cosmological rescaling can be applied directly to halo merger trees, allowing the Santa Cruz semi-analytic model to generate galaxy populations for many values of Omega_m and sigma_8 without new N-body runs for each. A single-parameter correction based on halo profiles removes most systematic mass bias in the rescaled trees. When these rescaled trees are used to train models for parameter estimation from the stellar mass function or two-point correlation function, 64 base simulations expanded to roughly 1000 samples perform as well as 750 dedicated simulations, and scaling to 3200 samples improves Omega_m recovery by about 25 percent. The entire rescaling step costs only 0.1 CPU hours per original simulation. This approach addresses the expense of densely sampling cosmological parameters for galaxy survey analyses.

Core claim

Extending cosmological rescaling to merger trees with a halo-profile-based correction controlled by one free parameter suppresses systematic bias in rescaled halo masses to below the percent level across the Omega_m-sigma_8 plane, enabling the Santa Cruz semi-analytic model to produce accurate galaxy summary statistics for new cosmologies at negligible added cost; as few as 64 base N-body simulations rescaled to about 1000 training samples match the accuracy of 750 dedicated N-body simulations for estimating Omega_m and sigma_8, with further improvement at 3200 realisations.

What carries the argument

Cosmological rescaling applied directly to merger trees together with a single-parameter halo-profile correction that adjusts masses to the target cosmology while feeding the Santa Cruz semi-analytic model.

If this is right

  • Galaxy summary statistics become available across a dense grid of Omega_m and sigma_8 without the cost of new N-body simulations for each point.
  • The computational overhead of rescaling all trees from one CAMELS-SAM simulation remains around 0.1 CPU hours.
  • Prediction accuracy for Omega_m increases by roughly 25 percent when the number of rescaled realisations rises from 1000 to 3200.
  • The same rescaled trees can be fed to other semi-analytic models at essentially zero extra cost.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The method could be combined with existing suites that already vary astrophysical parameters to produce joint cosmological and astrophysical training sets.
  • If the single-parameter correction proves stable, the approach might be tested on other N-body codes or initial condition generators beyond those used here.

Load-bearing premise

The single free parameter in the halo-profile correction removes bias in rescaled halo masses to below one percent without degrading the accuracy of the semi-analytic galaxy formation outputs.

What would settle it

Compare the stellar mass function and two-point correlation function obtained from the rescaled trees against those from a set of fully independent N-body simulations run at the same target Omega_m and sigma_8 values and check whether the differences stay below the accuracy levels reported for the 64-to-750 equivalence.

Figures

Figures reproduced from arXiv: 2606.10024 by Lucia A. Perez, Rachel S. Somerville, Raul E. Angulo, Richard Stiskalek, Sergio Contreras, Shy Genel.

Figure 1
Figure 1. Figure 1: Schematic of the merger tree rescaling from a high σ8 simulation (top) to a lower σ8 target cosmology (bottom). Circles represent halos at discrete snapshots, with arrows indicating the progenitor–descendant linkages. The main progenitor branch is highlighted in blue, while secondary branches (grey and green) merge into it at earlier times. Rescaling to lower σ8 (a cosmology with slower growth of structure… view at source ↗
Figure 2
Figure 2. Figure 2: Relative bias in halo virial mass Mvir, concentration c, and spin λBullock between individual rescaled and simulated halos for two cosmological steps: (i) σ8 = 0.8 → 0.85 at fixed Ωm = 0.3, and (ii) Ωm = 0.25 → 0.3 at fixed σ8 = 0.8, at the target redshifts z = 0.12 and z = 0.3, respectively. In the top right, we indicate the mean µ and standard deviation σ of the distributions. The distributions are consi… view at source ↗
Figure 3
Figure 3. Figure 3: Example evolution of virial mass, concentration, and spin along the main progenitor branch of a single matched halo for a cosmological step Ωm = 0.25 → 0.3 at fixed σ8 = 0.8. Solid blue lines show the evolution in the original cosmology, solid green lines show the same halo evolved in the target cosmology, and dotted red lines show the result of applying the rescaling procedure to the original halo. The re… view at source ↗
Figure 5
Figure 5. Figure 5: Bias in rescaled Mvir as a function of the target σ8 for steps from σ8 = 0.8 at fixed Ωm = 0.3, computed for halos with Mvir > 1012 h −1 M⊙. The bias increases with step size but remains modest, reaching only 1.8% for the largest step to σ8 = 1.0. Error bars represent 1σ bootstrap uncertainties. modest: even for the shift from σ8 = 0.8 to σ8 = 1.0, the bias is only 1.8%. 5.3. Galaxy population summary stat… view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of the HMFs for steps in σ8 (left; ˆz ′ = 0.12) and Ωm (right; ˆz ′ = 0.3). Top panels show the original cosmology at z = 0 (green), the target cosmology at z = ˆz ′ (blue), and the rescaled result at z = ˆz ′ (red); bottom panels show ratios to the target. For a step in σ8 at fixed Ωm, the HMFs already agree closely without rescaling, reflecting the near-equivalence to a time shift; rescaling m… view at source ↗
Figure 7
Figure 7. Figure 7: Comparison of the SMFs for a step Ωm = 0.25 → 0.3 at σ8 = 0.8. The top panel shows the original cosmology at z = 0 (green), the target cosmology at z = 0.3 (blue), and the rescaled result at z = 0.3 (red); the bottom panel shows the ratio of rescaled to target. The rescaled SMF, derived from rescaled merger trees processed with SC-SAM, closely matches the target. Error bars indicate 1σ Poisson uncertaintie… view at source ↗
Figure 8
Figure 8. Figure 8: Comparison of the galaxy real-space 2PCF for a step Ωm = 0.25 → 0.3 at σ8 = 0.8 for galaxies with M⋆ > 109 M⊙. The top panel shows the original cosmol￾ogy at z = 0 (green), the target cosmology at z = 0.3 (blue), and the rescaled result at z = 0.3 (red); the bottom panel shows the ratio of rescaled to target. The rescaled 2PCF closely matches the target, with deviations of order 5% on the smallest scales a… view at source ↗
Figure 9
Figure 9. Figure 9: PQMass validation of rescaled SMFs to test whether they are statistically consistent with SMFs constructed directly on top of N-body simulations in CAMELS-SAM. Each panel shows the distribution of χ 2 PQM values for Nrescaled = 1000 realisations generated from Nsims = 16, 25, 115, and 265 base simulations, compared to the expected χ 2 distribution with 49 degrees of freedom (red), corresponding to a partit… view at source ↗
Figure 10
Figure 10. Figure 10: RMSE in Ωm and σ8 as a function of the number of base N-body simulations. Blue points: models trained on a fixed set of 1000 rescaled simulations generated from a varying number of base simulations. Red points: models trained directly on N-body simulations without rescaling, where the training set size equals the number on the x-axis. The RMSE is evaluated on 150 held-out SAM galaxy catalogues from direct… view at source ↗
Figure 11
Figure 11. Figure 11: RMSE in Ωm and σ8 as a function of training-set size. Blue points: models trained on rescaled simulations generated from a fixed pool of 750 base N-body simulations, with the number of rescaled realisations varying along the x-axis. Red points: models trained directly on N-body simulations without rescaling. The RMSE is evaluated on 150 held-out SAM galaxy populations built on direct N-body simulations. F… view at source ↗
Figure 12
Figure 12. Figure 12: RMSE in Ωm and σ8 as a function of training-set size. Blue points: models trained on rescaled simulations generated from a fixed pool of only 25 base N-body simulations (10 from CAMELS-SAM plus 15 at σ8 = 1), with the number of rescaled realisations varying along the x-axis. Red points: models trained directly on N-body simulations without rescaling. The RMSE is evaluated on 150 held-out SAM galaxy popula… view at source ↗
Figure 13
Figure 13. Figure 13: Held-out RMSE in Ωm (left) and σ8 (right) as a function of the number NICs of unique initial-condition phases in the 79-simulation base set, shown as filled markers. Open markers show the reference runs of Section 6.2, in which all Nbase simulations have independent phases, so NICs = Nbase. Error bars give 1σ bootstrap uncertainties. The SMF RMSE is set entirely by rescaling error: cosmic variance makes n… view at source ↗
Figure 14
Figure 14. Figure 14: RMSE in Ωm and σ8 as a function of the number of base N-body simulations (x-axis) and rescaled realisations (y-axis), for models trained on the SMF (top) and 2PCF (bottom), evaluated on 150 held-out SAM galaxy catalogues derived from direct N-body simulations. Marker colour encodes the bootstrap mean RMSE; the divergent colour map is centred on the mean RMSE +1σ of the N-body baseline (trained on 750 simu… view at source ↗
Figure 15
Figure 15. Figure 15: RMSE in Ωm and σ8 as a function of training-set size. Blue points: models trained on rescaled simulations generated from a fixed pool of 79 base N-body simulations (64 from CAMELS-SAM plus 15 at σ8 = 1), with the number of rescaled realisations varying along the x-axis. Red points: models trained directly on N-body simulations without rescaling. The RMSE is evaluated on 150 held-out SAM galaxy populations… view at source ↗
read the original abstract

Learning cosmology from galaxy surveys requires large suites of simulations spanning the cosmological and astrophysical parameter space, yet hydrodynamical simulations of galaxy formation remain prohibitively expensive. Semi-analytic models offer an inexpensive, physically grounded alternative, but still require halo merger trees from $N$-body simulations, and densely sampling cosmological parameters in sufficient volume remains expensive. We address this by extending cosmological rescaling to operate directly on merger trees and applying it in the $\Omega_{\rm m}$-$\sigma_8$ plane, running the Santa Cruz semi-analytic model for galaxy formation on the rescaled trees to produce galaxy populations across new cosmological and astrophysical parameters at negligible additional cost. A novel halo-profile-based correction, controlled by a single free parameter, suppresses systematic bias in rescaled halo masses to below the per cent level. We apply the method to parameter estimation of $\Omega_{\rm m}$ and $\sigma_8$ given either the stellar mass function or the two-point correlation function, finding that as few as 64, and potentially fewer, base $N$-body simulations, rescaled to $\sim1000$ training samples, match the accuracy of 750 dedicated $N$-body simulations; rescaling to 3200 realisations improves the prediction of $\Omega_{\rm m}$ by $\sim25\%$. Rescaling all merger trees from a single CAMELS-SAM $N$-body simulation costs $\sim0.1$ CPUh, compared to several thousand CPUh to run the simulation itself. We demonstrate a practical route to obtaining predictions of galaxy summary statistics across cosmological and astrophysical parameters, even with a relatively small number of base $N$-body simulations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper extends cosmological rescaling techniques to operate directly on halo merger trees in the Ω_m-σ_8 plane. It applies a halo-profile-based correction controlled by a single free parameter to reduce systematic mass bias in the rescaled halos, then runs the Santa Cruz semi-analytic model on the corrected trees to generate galaxy populations. The central result is that rescaling from as few as 64 base N-body simulations to produce ~1000 training samples yields Ω_m and σ_8 inference accuracy comparable to that obtained from 750 dedicated N-body simulations when using either the stellar mass function or the two-point correlation function; further rescaling to 3200 realizations improves Ω_m prediction by ~25%. The computational cost of rescaling is stated to be negligible (~0.1 CPUh per CAMELS-SAM tree set).

Significance. If the rescaling plus single-parameter correction preserves merger-tree structure and SAM outputs to the claimed precision, the method offers a practical route to densely sampling cosmological and astrophysical parameter space for galaxy-survey inference at far lower cost than running equivalent suites of dedicated N-body simulations. This would be a useful efficiency gain for training sets used in cosmological parameter estimation from galaxy statistics.

major comments (2)
  1. [Section describing the halo-profile correction and its validation] The central accuracy claim (64 base simulations matching 750 dedicated ones) rests on the halo-profile correction suppressing mass bias to <1% uniformly across the Ω_m-σ_8 plane while leaving SAM galaxy statistics intact. The manuscript does not provide a quantitative validation (e.g., bias maps or residual distributions versus Ω_m and σ_8) demonstrating that a single scalar parameter achieves this uniformly; any cosmology-dependent residual from concentration-mass relations or subhalo stripping would propagate directly into the stellar-mass function and correlation function used for the inference comparison.
  2. [Results on Ω_m/σ_8 inference] The reported equivalence in parameter-estimation accuracy is presented without an explicit error budget or cross-validation that isolates the contribution of the rescaling correction from other modeling choices (e.g., the specific SAM parameter settings or the choice of summary statistics). It is therefore unclear whether the ~25% improvement quoted for 3200 realizations is robust to variations in the correction parameter or to the precise definition of the training-set size.
minor comments (2)
  1. [Methods] Notation for the single free parameter of the halo-profile correction should be introduced with an explicit equation and its fitting procedure described in a dedicated subsection.
  2. [Figure captions] Figure captions for any bias or accuracy plots should state the exact range of Ω_m and σ_8 values tested and whether the correction parameter was held fixed or re-tuned per cosmology.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed report. We address each major comment below and have revised the manuscript accordingly to strengthen the presentation of the halo-profile correction validation and the inference error budget.

read point-by-point responses

  1. Referee: [Section describing the halo-profile correction and its validation] The central accuracy claim (64 base simulations matching 750 dedicated ones) rests on the halo-profile correction suppressing mass bias to <1% uniformly across the Ω_m-σ_8 plane while leaving SAM galaxy statistics intact. The manuscript does not provide a quantitative validation (e.g., bias maps or residual distributions versus Ω_m and σ_8) demonstrating that a single scalar parameter achieves this uniformly; any cosmology-dependent residual from concentration-mass relations or subhalo stripping would propagate directly into the stellar-mass function and correlation function used for the inference comparison.

    Authors: We agree that explicit quantitative validation maps would strengthen the manuscript. In the revised version we will add bias maps and residual distributions of the corrected halo masses versus Ω_m and σ_8, confirming uniformity to <1% across the plane. The single-parameter correction is applied at the halo-profile level precisely to absorb cosmology-dependent effects from concentration-mass relations; any residual subhalo-stripping mismatches would appear as systematic offsets in the SAM outputs, yet the reported SMF and 2PCF inference accuracies match those from dedicated runs to within the quoted precision. We will add a short discussion of this point. revision: yes

  2. Referee: [Results on Ω_m/σ_8 inference] The reported equivalence in parameter-estimation accuracy is presented without an explicit error budget or cross-validation that isolates the contribution of the rescaling correction from other modeling choices (e.g., the specific SAM parameter settings or the choice of summary statistics). It is therefore unclear whether the ~25% improvement quoted for 3200 realizations is robust to variations in the correction parameter or to the precise definition of the training-set size.

    Authors: The equivalence is demonstrated by direct side-by-side comparison of the recovered Ω_m–σ_8 posteriors. To isolate the rescaling contribution we will add, in revision, an explicit error-budget subsection together with cross-validation tests that vary the correction parameter by ±20% around its fiducial value and recompute the inference accuracy. The training-set size is defined as the number of rescaled merger-tree realizations supplied to the emulator; we will clarify this definition and show the scaling of accuracy with training-set size (including the 1000-to-3200 step) to confirm robustness of the ~25% Ω_m improvement. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's derivation introduces a novel halo-profile correction with one explicit free parameter to suppress rescaled halo mass bias, then validates the resulting SAM galaxy statistics against dedicated N-body runs. No step reduces by construction to its own inputs: the parameter is fitted to halo masses but the target claims (matching accuracy for Ω_m/σ_8 inference) rest on external comparisons, not self-definition or renaming. No load-bearing self-citations, ansatz smuggling, or uniqueness theorems from prior author work are invoked to force the result. The chain is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The ledger records the single free parameter for the correction and the domain assumption that rescaling preserves merger tree properties needed by the SAM.

free parameters (1)
  • halo-profile correction parameter
    Single free parameter that controls suppression of systematic bias in rescaled halo masses to the per cent level.
axioms (1)
  • domain assumption Cosmological rescaling applied directly to merger trees preserves the statistical properties required for accurate semi-analytic galaxy formation modeling in the Ω_m-σ_8 plane
    This assumption underpins the entire extension of the rescaling technique.

pith-pipeline@v0.9.1-grok · 5858 in / 1297 out tokens · 28272 ms · 2026-06-27T15:18:00.435869+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

96 extracted references · 90 canonical work pages · 40 internal anchors

  1. [1]

    Cosmological constraints from the small scale clustering of Emission Line Galaxies

    Cosmological constraints from the small scale clustering of Emission Line Galaxies. arXiv e-prints , keywords =. doi:10.48550/arXiv.2604.19449 , archivePrefix =. 2604.19449 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2604.19449

  2. [2]

    , keywords =

    Cosmological constraints from galaxy clustering and galaxy─galaxy lensing with extended SubHalo Abundance Matching. , keywords =. doi:10.1093/mnras/staf2143 , archivePrefix =. 2507.01601 , primaryClass =

    work page doi:10.1093/mnras/staf2143

  3. [3]

    , keywords =

    Non-universality of the mass function: dependence on the growth rate and power spectrum shape. , keywords =. doi:10.1093/mnras/stab3337 , archivePrefix =. 2102.08958 , primaryClass =

    work page doi:10.1093/mnras/stab3337

  4. [4]

    , keywords =

    The CAMELS Project: Cosmology and Astrophysics with Machine-learning Simulations. , keywords =. doi:10.3847/1538-4357/abf7ba , archivePrefix =. 2010.00619 , primaryClass =

    work page doi:10.3847/1538-4357/abf7ba 2010

  5. [5]

    , keywords =

    Constraining Cosmology with Machine Learning and Galaxy Clustering: The CAMELS-SAM Suite. , keywords =. doi:10.3847/1538-4357/accd52 , archivePrefix =. 2204.02408 , primaryClass =

    work page doi:10.3847/1538-4357/accd52

  6. [6]

    , keywords =

    The AREPO Public Code Release. , keywords =. doi:10.3847/1538-4365/ab908c , archivePrefix =. 1909.04667 , primaryClass =

    work page doi:10.3847/1538-4365/ab908c 1909

  7. [7]

    CAMB: Code for Anisotropies in the Microwave Background

  8. [8]

    The Rockstar Phase-Space Temporal Halo Finder and the Velocity Offsets of Cluster Cores

    The ROCKSTAR Phase-space Temporal Halo Finder and the Velocity Offsets of Cluster Cores. , keywords =. doi:10.1088/0004-637X/762/2/109 , archivePrefix =. 1110.4372 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1088/0004-637x/762/2/109

  9. [9]

    Gravitationally Consistent Halo Catalogs and Merger Trees for Precision Cosmology

    Gravitationally Consistent Halo Catalogs and Merger Trees for Precision Cosmology. , keywords =. doi:10.1088/0004-637X/763/1/18 , archivePrefix =. 1110.4370 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1088/0004-637x/763/1/18

  10. [10]

    , keywords =

    Semi-analytic modelling of galaxy formation: the local Universe. , keywords =. doi:10.1046/j.1365-8711.1999.03032.x , archivePrefix =. astro-ph/9802268 , primaryClass =

    work page doi:10.1046/j.1365-8711.1999.03032.x 1999

  11. [11]

    2009, MNRAS, 394, 191, doi: 10.1111/j.1365-2966.2008.14359.x

    A semi-analytic model for the co-evolution of galaxies, black holes and active galactic nuclei. , keywords =. doi:10.1111/j.1365-2966.2008.13805.x , archivePrefix =. 0808.1227 , primaryClass =

    work page doi:10.1111/j.1365-2966.2008.13805.x 2008

  12. [12]

    Star Formation in Semi-Analytic Galaxy Formation Models with Multiphase Gas

    Star formation in semi-analytic galaxy formation models with multiphase gas. , keywords =. doi:10.1093/mnras/stv1877 , archivePrefix =. 1503.00755 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stv1877

  13. [13]

    , keywords =

    Mock light-cones and theory friendly catalogues for the CANDELS survey. , keywords =. doi:10.1093/mnras/stab231 , archivePrefix =. 2102.00108 , primaryClass =

    work page doi:10.1093/mnras/stab231

  14. [14]

    Galaxy scaling relations, dispersions, and residuals at z = 0

    Galaxy formation in the Santa Cruz semi-analytic model compared with IllustrisTNG - I. Galaxy scaling relations, dispersions, and residuals at z = 0. , keywords =. doi:10.1093/mnras/stac2297 , archivePrefix =. 2111.03077 , primaryClass =

    work page doi:10.1093/mnras/stac2297

  15. [15]

    2010, Mon

    One simulation to fit them all - changing the background parameters of a cosmological N-body simulation. , keywords =. doi:10.1111/j.1365-2966.2010.16459.x , archivePrefix =. 0912.4277 , primaryClass =

    work page doi:10.1111/j.1365-2966.2010.16459.x 2010

  16. [16]

    , keywords =

    How accurate is it to update the cosmology of your halo catalogues?. , keywords =. doi:10.1111/j.1365-2966.2011.19635.x , archivePrefix =. 1103.5074 , primaryClass =

    work page doi:10.1111/j.1365-2966.2011.19635.x 2011

  17. [17]

    Galaxy formation in WMAP1 and WMAP7 cosmologies

    Galaxy formation in WMAP1 and WMAP7 cosmologies. , keywords =. doi:10.1093/mnras/sts115 , archivePrefix =. 1206.0052 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/sts115

  18. [18]

    Remapping dark matter halo catalogues between cosmological simulations

    Remapping dark matter halo catalogues between cosmological simulations. , keywords =. doi:10.1093/mnras/stu345 , archivePrefix =. 1308.5183 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stu345

  19. [19]

    A Universal Density Profile from Hierarchical Clustering

    A Universal Density Profile from Hierarchical Clustering. , keywords =. doi:10.1086/304888 , archivePrefix =. astro-ph/9611107 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/304888

  20. [20]

    Statistical Properties of X-ray Clusters: Analytic and Numerical Comparisons

    Statistical Properties of X-Ray Clusters: Analytic and Numerical Comparisons. , keywords =. doi:10.1086/305262 , archivePrefix =. astro-ph/9710107 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/305262

  21. [21]

    NIHAO project II: Halo shape, phase-space density and velocity distribution of dark matter in galaxy formation simulations

    NIHAO project II: halo shape, phase-space density and velocity distribution of dark matter in galaxy formation simulations. , keywords =. doi:10.1093/mnras/stw1688 , archivePrefix =. 1503.04814 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stw1688

  22. [22]

    On the galaxy-halo connection in the EAGLE simulation

    On the galaxy-halo connection in the EAGLE simulation. , keywords =. doi:10.1093/mnrasl/slx093 , archivePrefix =. 1612.01029 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnrasl/slx093

  23. [23]

    Comparing galaxy formation in semi-analytic models and hydrodynamical simulations

    Comparing galaxy formation in semi-analytic models and hydrodynamical simulations. , keywords =. doi:10.1093/mnras/stx2770 , archivePrefix =. 1709.08647 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stx2770

  24. [24]

    , keywords =

    Baryons shaping dark matter haloes. , keywords =. doi:10.1093/mnras/staa3988 , archivePrefix =. 2008.02404 , primaryClass =

    work page doi:10.1093/mnras/staa3988 2008

  25. [25]

    , keywords =

    Evaluating the variance of individual halo properties in constrained cosmological simulations. , keywords =. doi:10.1093/mnras/stae2292 , archivePrefix =. 2310.20672 , primaryClass =

    work page doi:10.1093/mnras/stae2292

  26. [26]

    The Mass-Concentration-Redshift Relation of Cold and Warm Dark Matter Halos

    The mass-concentration-redshift relation of cold and warm dark matter haloes. , keywords =. doi:10.1093/mnras/stw1046 , archivePrefix =. 1601.02624 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stw1046

  27. [27]

    , keywords =

    CORRFUNC - a suite of blazing fast correlation functions on the CPU. , keywords =. doi:10.1093/mnras/stz3157 , archivePrefix =. 1911.03545 , primaryClass =

    work page doi:10.1093/mnras/stz3157 1911

  28. [28]

    D., & Szalay, A

    Bias and Variance of Angular Correlation Functions. , keywords =. doi:10.1086/172900 , adsurl =

    work page doi:10.1086/172900

  29. [29]

    Fast $\epsilon$-free Inference of Simulation Models with Bayesian Conditional Density Estimation

    Fast -free Inference of Simulation Models with Bayesian Conditional Density Estimation. arXiv e-prints , keywords =. doi:10.48550/arXiv.1605.06376 , archivePrefix =. 1605.06376 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1605.06376

  30. [30]

    Approximate Bayesian Computational methods

    Approximate Bayesian Computational methods. arXiv e-prints , keywords =. doi:10.48550/arXiv.1101.0955 , archivePrefix =. 1101.0955 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1101.0955

  31. [31]

    Proceedings of the National Academy of Science , keywords =

    The frontier of simulation-based inference. Proceedings of the National Academy of Science , keywords =. doi:10.1073/pnas.1912789117 , archivePrefix =. 1911.01429 , primaryClass =

    work page doi:10.1073/pnas.1912789117 1911

  32. [32]

    Automatic Posterior Transformation for Likelihood-Free Inference

    Automatic Posterior Transformation for Likelihood-Free Inference. arXiv e-prints , keywords =. doi:10.48550/arXiv.1905.07488 , archivePrefix =. 1905.07488 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1905.07488 1905

  33. [33]

    Massive optimal data compression and density estimation for scalable, likelihood-free inference in cosmology

    Massive optimal data compression and density estimation for scalable, likelihood-free inference in cosmology. , keywords =. doi:10.1093/mnras/sty819 , archivePrefix =. 1801.01497 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/sty819

  34. [34]

    Sequential Neural Likelihood: Fast Likelihood-free Inference with Autoregressive Flows

    Sequential Neural Likelihood: Fast Likelihood-free Inference with Autoregressive Flows. arXiv e-prints , keywords =. doi:10.48550/arXiv.1805.07226 , archivePrefix =. 1805.07226 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1805.07226

  35. [35]

    arXiv e-prints , keywords =

    Likelihood-free MCMC with Amortized Approximate Ratio Estimators. arXiv e-prints , keywords =. doi:10.48550/arXiv.1903.04057 , archivePrefix =. 1903.04057 , primaryClass =

    work page doi:10.48550/arxiv.1903.04057 1903

  36. [36]

    arXiv e-prints , keywords =

    Normalizing Flows for Probabilistic Modeling and Inference. arXiv e-prints , keywords =. doi:10.48550/arXiv.1912.02762 , archivePrefix =. 1912.02762 , primaryClass =

    work page doi:10.48550/arxiv.1912.02762 1912

  37. [37]

    Mixture density networks

    Bishop, Christopher M. Mixture density networks. 1994

    1994

  38. [38]

    Optuna: A Next-generation Hyperparameter Optimization Framework

    Optuna: A Next-generation Hyperparameter Optimization Framework. arXiv e-prints , keywords =. doi:10.48550/arXiv.1907.10902 , archivePrefix =. 1907.10902 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1907.10902 1907

  39. [39]

    The Open Journal of Astrophysics , keywords =

    LtU-ILI: An All-in-One Framework for Implicit Inference in Astrophysics and Cosmology. The Open Journal of Astrophysics , keywords =. doi:10.33232/001c.120559 , archivePrefix =. 2402.05137 , primaryClass =

    work page doi:10.33232/001c.120559

  40. [40]

    Cook and Andrew Gelman and Donald B

    Samantha R. Cook and Andrew Gelman and Donald B. Rubin , journal =. Validation of Software for Bayesian Models Using Posterior Quantiles , urldate =

  41. [41]

    arXiv e-prints , keywords =

    Diagnostics for Conditional Density Models and Bayesian Inference Algorithms. arXiv e-prints , keywords =. doi:10.48550/arXiv.2102.10473 , archivePrefix =. 2102.10473 , primaryClass =

    work page doi:10.48550/arxiv.2102.10473

  42. [42]

    40th International Conference on Machine Learning , keywords =

    Sampling-Based Accuracy Testing of Posterior Estimators for General Inference. 40th International Conference on Machine Learning , keywords =. doi:10.48550/arXiv.2302.03026 , archivePrefix =. 2302.03026 , primaryClass =

    work page doi:10.48550/arxiv.2302.03026

  43. [43]

    Gaussian Error Linear Units (GELUs)

    Gaussian Error Linear Units (GELUs). arXiv e-prints , keywords =. doi:10.48550/arXiv.1606.08415 , archivePrefix =. 1606.08415 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1606.08415

  44. [44]

    , keywords =

    The BACCO simulation project: exploiting the full power of large-scale structure for cosmology. , keywords =. doi:10.1093/mnras/stab2018 , archivePrefix =. 2004.06245 , primaryClass =

    work page doi:10.1093/mnras/stab2018 2004

  45. [45]

    A Quantification of the Butterfly Effect in Cosmological Simulations and Implications for Galaxy Scaling Relations

    A Quantification of the Butterfly Effect in Cosmological Simulations and Implications for Galaxy Scaling Relations. , keywords =. doi:10.3847/1538-4357/aaf4bb , archivePrefix =. 1807.07084 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4357/aaf4bb

  46. [46]

    , keywords =

    3 per cent-accurate predictions for the clustering of dark matter, haloes, and subhaloes, over a wide range of cosmologies and scales. , keywords =. doi:10.1093/mnras/staa3117 , archivePrefix =. 2001.03176 , primaryClass =

    work page doi:10.1093/mnras/staa3117 2001

  47. [47]

    A universal angular momentum profile for galactic halos

    A Universal Angular Momentum Profile for Galactic Halos. , keywords =. doi:10.1086/321477 , archivePrefix =. astro-ph/0011001 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/321477

  48. [48]

    Halo and Subhalo Demographics with Planck Cosmological Parameters: Bolshoi-Planck and MultiDark-Planck Simulations

    Halo and subhalo demographics with Planck cosmological parameters: Bolshoi-Planck and MultiDark-Planck simulations. , keywords =. doi:10.1093/mnras/stw1705 , archivePrefix =. 1602.04813 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stw1705

  49. [49]

    The effects of Non-Gaussian initial conditions on the structure and substructure of Cold Dark Matter halos

    The Effects of Non-Gaussian Initial Conditions on the Structure and Substructure of Cold Dark Matter Halos. , keywords =. doi:10.1086/378773 , archivePrefix =. astro-ph/0306293 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/378773

  50. [50]

    The large-scale structure of the universe

  51. [51]

    , keywords =

    Cosmological remapping for efficient generation of 21 cm intensity mapping mocks. , keywords =. doi:10.1088/1475-7516/2026/01/019 , archivePrefix =. 2506.14588 , primaryClass =

    work page doi:10.1088/1475-7516/2026/01/019 2026

  52. [52]

    Masked Autoregressive Flow for Density Estimation

    Masked Autoregressive Flow for Density Estimation. arXiv e-prints , keywords =. doi:10.48550/arXiv.1705.07057 , archivePrefix =. 1705.07057 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1705.07057

  53. [53]

    Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning

    Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning. arXiv e-prints , keywords =. doi:10.48550/arXiv.1506.02142 , archivePrefix =. 1506.02142 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1506.02142

  54. [54]

    SGDR: Stochastic Gradient Descent with Warm Restarts

    SGDR: Stochastic Gradient Descent with Warm Restarts. arXiv e-prints , keywords =. doi:10.48550/arXiv.1608.03983 , archivePrefix =. 1608.03983 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1608.03983

  55. [55]

    , keywords =

    The BACCO simulation project: a baryonification emulator with neural networks. , keywords =. doi:10.1093/mnras/stab1911 , archivePrefix =. 2011.15018 , primaryClass =

    work page doi:10.1093/mnras/stab1911 2011

  56. [56]

    , keywords =

    The BACCO simulation project: biased tracers in real space. , keywords =. doi:10.1093/mnras/stad2008 , archivePrefix =. 2101.12187 , primaryClass =

    work page doi:10.1093/mnras/stad2008

  57. [57]

    , keywords =

    The bacco simulation project: bacco hybrid Lagrangian bias expansion model in redshift space. , keywords =. doi:10.1093/mnras/stad368 , archivePrefix =. 2207.06437 , primaryClass =

    work page doi:10.1093/mnras/stad368

  58. [58]

    The Thirteenth International Conference on Learning Representations , year = 2025, month = jan, eid =

    PQMass: Probabilistic Assessment of the Quality of Generative Models using Probability Mass Estimation. The Thirteenth International Conference on Learning Representations , year = 2025, month = jan, eid =. doi:10.48550/arXiv.2402.04355 , archivePrefix =. 2402.04355 , primaryClass =

    work page doi:10.48550/arxiv.2402.04355 2025

  59. [59]

    , keywords =

    The BIG SOBOL SEQUENCE: How many simulations do we need for simulation-based inference in cosmology?. , keywords =. doi:10.1051/0004-6361/202554602 , archivePrefix =. 2503.13755 , primaryClass =

    work page doi:10.1051/0004-6361/202554602

  60. [60]

    Spectroscopic Target Selection in the Sloan Digital Sky Survey: The Main Galaxy Sample

    Spectroscopic Target Selection in the Sloan Digital Sky Survey: The Main Galaxy Sample. , keywords =. doi:10.1086/342343 , archivePrefix =. astro-ph/0206225 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1086/342343

  61. [61]

    The Dark Energy Spectroscopic Instrument (DESI)

    The Dark Energy Spectroscopic Instrument (DESI). Bulletin of the American Astronomical Society , year = 2019, volume =. doi:10.48550/arXiv.1907.10688 , archivePrefix =. 1907.10688 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1907.10688 2019

  62. [62]

    Overview of the Euclid mission

    Euclid: I. Overview of the Euclid mission. , keywords =. doi:10.1051/0004-6361/202450810 , archivePrefix =. 2405.13491 , primaryClass =

    work page doi:10.1051/0004-6361/202450810

  63. [63]

    LSST: from Science Drivers to Reference Design and Anticipated Data Products

    LSST: From Science Drivers to Reference Design and Anticipated Data Products. , keywords =. doi:10.3847/1538-4357/ab042c , archivePrefix =. 0805.2366 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4357/ab042c

  64. [64]

    Planck 2018 results. VI. Cosmological parameters. , keywords =. doi:10.1051/0004-6361/201833910 , archivePrefix =. 1807.06209 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1051/0004-6361/201833910 2018

  65. [65]

    2014 , volume =

    The No-U-Turn Sampler: Adaptively Setting Path Lengths in Hamiltonian Monte Carlo , journal =. 2014 , volume =

    2014

  66. [66]

    , keywords =

    Learning the Universe: cosmological and astrophysical parameter inference with galaxy luminosity functions and colours. , keywords =. doi:10.1093/mnras/staf1888 , archivePrefix =. 2411.13960 , primaryClass =

    work page doi:10.1093/mnras/staf1888

  67. [67]

    First results from the IllustrisTNG simulations: matter and galaxy clustering

    First results from the IllustrisTNG simulations: matter and galaxy clustering. , keywords =. doi:10.1093/mnras/stx3304 , archivePrefix =. 1707.03397 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stx3304

  68. [68]

    Simulating Galaxy Formation with the IllustrisTNG Model

    Simulating galaxy formation with the IllustrisTNG model. , keywords =. doi:10.1093/mnras/stx2656 , archivePrefix =. 1703.02970 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stx2656

  69. [69]

    The EAGLE project: Simulating the evolution and assembly of galaxies and their environments

    The EAGLE project: simulating the evolution and assembly of galaxies and their environments. , keywords =. doi:10.1093/mnras/stu2058 , archivePrefix =. 1407.7040 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stu2058

  70. [70]

    Simba: Cosmological Simulations with Black Hole Growth and Feedback

    SIMBA: Cosmological simulations with black hole growth and feedback. , keywords =. doi:10.1093/mnras/stz937 , archivePrefix =. 1901.10203 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stz937 1901

  71. [71]

    , keywords =

    The Quijote Simulations. , keywords =. doi:10.3847/1538-4365/ab9d82 , archivePrefix =. 1909.05273 , primaryClass =

    work page doi:10.3847/1538-4365/ab9d82 1909

  72. [72]

    Dancing in the dark: galactic properties trace spin swings along the cosmic web

    Dancing in the dark: galactic properties trace spin swings along the cosmic web. , keywords =. doi:10.1093/mnras/stu1227 , archivePrefix =. 1402.1165 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stu1227

  73. [73]

    Fast likelihood-free cosmology with neural density estimators and active learning

    Fast likelihood-free cosmology with neural density estimators and active learning. , keywords =. doi:10.1093/mnras/stz1960 , archivePrefix =. 1903.00007 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1093/mnras/stz1960 1903

  74. [74]

    , keywords =

    On the accuracy of dark matter halo merger trees and the consequences for semi-analytic models of galaxy formation. , keywords =. doi:10.1093/mnras/staf519 , archivePrefix =. 2501.07677 , primaryClass =

    work page doi:10.1093/mnras/staf519

  75. [75]

    Achieving Convergence in Galaxy Formation Models by Augmenting N-body Merger Trees

    Achieving convergence in galaxy formation models by augmenting N-body merger trees. Computational Astrophysics and Cosmology , keywords =. doi:10.1186/s40668-016-0016-3 , archivePrefix =. 1604.02147 , primaryClass =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1186/s40668-016-0016-3

  76. [76]

    , keywords =

    Transfer learning for multifidelity simulation-based inference in cosmology. , keywords =. doi:10.1093/mnras/staf1436 , archivePrefix =. 2505.21215 , primaryClass =

    work page doi:10.1093/mnras/staf1436

  77. [77]

    , keywords =

    Cosmological constraints using simulation-based inference of galaxy clustering with marked power spectra. , keywords =. doi:10.1103/9l4n-3pcx , archivePrefix =. 2404.04228 , primaryClass =

    work page doi:10.1103/9l4n-3pcx

  78. [78]

    , keywords =

    Galaxy Formation through Hierarchical Clustering. , year = 1991, month = sep, volume =. doi:10.1086/170483 , adsurl =

    work page doi:10.1086/170483 1991

  79. [79]

    , year = 1993, month = sep, volume =

    The formation and evolution of galaxies within merging dark matter haloes. , year = 1993, month = sep, volume =. doi:10.1093/mnras/264.1.201 , adsurl =

    work page doi:10.1093/mnras/264.1.201 1993

  80. [80]

    R., Pols, O

    Hierarchical galaxy formation. , keywords =. doi:10.1046/j.1365-8711.2000.03879.x , archivePrefix =. astro-ph/0007281 , primaryClass =

    work page doi:10.1046/j.1365-8711.2000.03879.x 2000

Showing first 80 references.