Implicit Likelihood Inference and $z$-Binned Reconstruction of Dark Energy $w(z)$

Jianbo Lu; Jia-Yi Feng; Ke Wang

REVIEW 3 major objections 4 minor 41 references

Implicit Likelihood Inference and $z$-Binned Reconstruction of Dark Energy $w(z)$

T0 review · 3 major / 4 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 The paper reconstructs the dark energy equation of state with a seven-bin piecewise model using simulation-based inference, finding $w_0 = -0.90 \pm 0.05$ in the lowest bin while higher bins remain consistent with the cosmological constant.

desk verdict Credible SBI demonstration for z-binned w(z), but the w0 dynamical-DE claim rests on a miscalibrated marginal and is not supported. read the letter →

arxiv 2608.08007 v1 pith:4EIYDZGJ submitted 2026-08-08 astro-ph.CO

Ke Wang , Jia-Yi Feng , Jianbo Lu This is my paper · ORCID

classification astro-ph.CO PACS 95.36.+x98.80.Es

keywords darkenergyequationofstateredshiftbinningsimulation-basedinferenceneurallikelihoodestimationcosmologicalparametersCMBpowerspectraBAOdistanceratios

topics Dark Energy

open problems Dark Energy

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 sets out to reconstruct the dark energy equation-of-state parameter $w(z)$ without assuming a functional form: it slices redshift into seven bins, treats $w$ as constant inside each bin, and fits all bins together with the six base cosmological parameters using simulation-based inference. The data are the Planck 2018 CMB temperature and polarization spectra with lensing, the DESI DR2 BAO distance ratios, and the Pantheon+ supernova apparent magnitudes, brought together through forward simulators built on a Boltzmann solver. The central result is that the lowest-redshift bin prefers $w_0 = -0.90 \pm 0.05$, about two standard deviations away from the cosmological-constant value $w = -1$, while the other constrained bins agree with $-1$ at 68% confidence. A sympathetic reader would take this as evidence that current data may already be hinting at dynamical dark energy at recent times, and that implicit likelihood methods can handle a piecewise $w(z)$ reconstruction that explicit likelihood analyses struggle with.

What carries the argument

The carrying object is the $w_i$CDM model, defined by a piecewise-constant equation of state $w(z) = w_i$ on $z_i < z \le z_{i+1}$ with bins $z_i = \{0, 0.4, 0.6, 0.8, 1.1, 1.4, 1.9, \infty\}$, so that the dark-energy density evolves continuously across bin boundaries even though $w$ jumps. The inference machinery is Sequential Neural Likelihood Estimation (SNLE), which trains a neural density estimator (a masked autoregressive flow) to approximate the likelihood $q_W(x|\theta)$ from simulated data-parameter pairs, then uses it to sample the posterior for the observed data. Six sequential rounds of 20,000 simulations each refine the learned likelihood, and validation with rank statistics and percentile-percentile plots checks that the approximate posterior is calibrated.

What would settle it

Run a dedicated marginal coverage test on $w_0$ alone with a large test set (e.g., 10,000 simulations) and compare the empirical coverage of the reported 68% credible interval; if the interval covers the true $w_0$ less than, say, 60% of the time, the $2\sigma$ deviation is not trustworthy. Equivalently, an explicit-likelihood MCMC analysis of the same data could be checked to see whether it also finds $w_0 \approx -0.90$.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a reconstruction of $w(z)$ from a $w_i$CDM model with seven piecewise-constant bins, obtained by sequentially training a neural likelihood estimator over six rounds of 20,000 forward simulations. The learned posterior, validated on 1,000 held-out simulations with rank statistics and percentile-percentile plots, gives 68% constraints $w_0 = -0.90 \pm 0.05$, $w_1 = -0.75 \pm 0.37$, $w_2 = -1.07^{+0.82}_{-0.86}$, $w_3 = -1.90^{+1.14}_{-1.04}$, $w_4 = -0.19^{+1.27}_{-1.32}$, while $w_5$ and $w_6$ remain essentially unconstrained. The reconstruction marginally favors dynamical dark energy in the first redshift bin and is consistent with the cosmological constant at 68% C.L. in the other bins.

Load-bearing premise

The load-bearing premise is that the learned neural likelihood is correctly calibrated for $w_0$, so the roughly $2\sigma$ preference for $w_0 > -1$ describes the data rather than the degeneracy ridge between $H_0$ and $w_0$.

Editorial extensions

If this is right

The seven-bin reconstruction demonstrates that $w(z)$ can be mapped without parametric priors or theoretical bin-prior covariances, because the simulation-based pipeline fits the CMB power spectra globally at this dimensionality.
The low-redshift preference for $w_0 > -1$ by roughly $2\sigma$, if taken at face value, indicates that dark energy behaved differently from a cosmological constant at $z < 0.4$.
The fact that $w_1$ through $w_4$ stay within 68% of $-1$ means the DESI BAO and Pantheon+ data do not yet require dynamics beyond the first bin.
The unconstrained $w_5$ and $w_6$ are a data-sparsity statement: additional BAO or supernova data at $z \gtrsim 1$ can be added without greatly increasing computational cost, unlike explicit likelihood approaches.

Reading between the lines

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

The paper attributes the rank-statistic spike for $w_0$ and $H_0$ to their anti-correlation; a dedicated marginal coverage test for $w_0$ alone would determine whether this spike is a harmless degeneracy or a calibration failure that could soften the $2\sigma$ claim.
If future data sharpen the first-bin deviation, the same locally amortized posterior could be used to test whether the transition happens at a specific redshift or evolves smoothly, because new observations can be folded in without retraining from scratch.
The method transfers naturally to other high-dimensional cosmological likelihoods, such as joint constraints on neutrino mass, curvature, and dark energy, where explicit likelihoods would be impractical.

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

Fairly solid application paper with one load-bearing weakness. The genuinely new thing here is putting the LtU-ILI/SNLE pipeline to work on a 7-bin piecewise-constant w(z) model with full Planck CMB spectra, DESI DR2, and Pantheon+, sampling 14 parameters jointly. They give the simulators, training setup, and three validation diagnostics. The w2–w4 bins come out consistent with -1 and the posterior broadening at high z is honestly reported. That part is worth a look for anyone doing SBI in cosmology.

The problem is the w0 result. They report w0 = -0.90 ± 0.05, a ~2σ deviation, and 'marginally favors dynamical DE.' The only support is the learned marginal posterior of w0. Their own rank statistic for w0 shows a boundary spike. The paper says this is due to the H0–w0 anti-correlation, but that doesn't work: rank statistics are computed per parameter after marginalization, so a correlation between H0 and w0 cannot make the w0 ranks nonuniform unless the learned w0 marginal itself is wrong. A boundary spike means the true w0 often falls outside the learned posterior, which is exactly the kind of miscalibration that would corrupt the 68% interval. So the ~2σ claim is not established. They need a dedicated marginal coverage test for w0 alone and a cross-check against a standard explicit-likelihood MCMC analysis.

Two smaller issues. First, no look-elsewhere correction across the seven bins; with five bins consistent with -1, the chance that one bin shows ~2σ is not negligible. Second, no code or data release, so the calibration check can't be reproduced. Also, the 'novel method' framing is overstated—the pipeline and the wiCDM model come from cited prior work; the application is new, not the machinery.

If the w0 calibration issue is fixed and the claim survives, this becomes a clean demonstration. As it stands, I'd advise a referee to demand the coverage test and the explicit-likelihood comparison before accepting 'marginally favors dynamical DE.' The rest of the paper is fine.

Referee Report

3 major / 4 minor

Summary. The paper proposes a simulation-based inference (SBI) approach to reconstruct the dark-energy equation of state w(z) as a piecewise-constant function in seven redshift bins within a w_iCDM model. The authors build forward simulators for Planck 2018 CMB power spectra and lensing, DESI DR2 BAO distance ratios, and Pantheon+ supernova apparent magnitudes using CLASS, and train a neural likelihood estimator via the LtU-ILI pipeline with 6 rounds of 20,000 simulations each. After validation on test simulations using rank statistics and P-P plots, they apply the learned posterior to the observed data and report constraints on the 14 parameters (Table I). The headline result is that w0 = -0.90 ± 0.05, which is about 2σ above -1, and they conclude that the reconstruction 'marginally favors dynamical DE in the first bin' while other constrained bins are consistent with ΛCDM at 68% C.L.

Significance. If the result is correct, the work is methodologically interesting: it demonstrates a multi-dimensional SBI analysis of a 14-parameter w_iCDM model with global CMB fits, a regime where traditional explicit-likelihood sampling scales poorly. The code is built on public simulators (CLASS, sbi, emcee), and the paper provides training details, priors, and validation diagnostics. However, the central scientific claim — a ~2σ preference for w0 > -1 — rests on the calibratedness of the learned marginal posterior for w0, and the paper's own validation shows a rank-statistic spike for w0 that is not convincingly explained. The claim is therefore not yet established.

major comments (3)

[§III, Fig. 2 and Eq. (19)] The rank-statistic spike for w0 is direct evidence that the learned marginal posterior for w0 is miscalibrated. The explanation that the spike 'results from the anti-correlation between H0 and w0' is not valid for a marginal rank statistic: the rank of the true w0 is computed among posterior samples of w0 alone, after H0 has been marginalized out. A correlation between H0 and w0 in the joint posterior cannot induce nonuniformity in the w0 marginal rank unless the learned marginal itself deviates from the true marginal. A boundary spike indicates that the true w0 often falls outside the learned posterior support, meaning the 68% interval reported in Table I for w0 may be too narrow or shifted. Since the claim of dynamical dark energy rests on w0 = -0.90 ± 0.05 being about 2σ from -1, this miscalibration directly undermines the headline result.
[§III, Table I and Fig. 3] The paper does not provide a dedicated marginal coverage test for w0 alone, nor does it quantify how much of the rank-statistic spike is driven by the H0-w0 ridge. The P-P plot in Fig. 3 shows disagreement for w0 (as acknowledged in the text), but the magnitude and location of the miscalibration are not quantified. A simulation-based calibration test restricted to the w0 marginal, or a comparison of the learned posterior against an explicit-likelihood analysis of the same data (e.g., using MontePython or Cobaya with an equivalent w_iCDM model), would establish whether the reported 68% interval is credible. Without such a check, the conclusion that the data 'marginally favor dynamical DE in the first bin' is unsupported.
[§III and §II A] The validation tests are performed on test simulations drawn from the same forward model, which checks internal consistency of the inference pipeline but not the fidelity of the forward model. In particular, the CMB noise realizations at ℓ>52 are generated from a covariance matrix fixed by a fiducial ΛCDM cosmology (Eq. 10), and the observed likelihood is assumed to be represented by these noise realizations. Systematic errors in this forward model would not be detected by the rank statistics. I recommend adding an explicit-likelihood cross-check on the actual observed data to verify that the posterior is consistent with established cosmological constraints, especially for the six base parameters that are well measured by Planck.

minor comments (4)

[Abstract and Introduction] There are several typographical and rendering issues in the manuscript text, including 'z' instead of 'z' in the title and abstract, and the garbled phrases 'smi. . .ndclpp p teb consext8' in the description of the Planck lensing covariance. These should be corrected in a revision.
[Eq. (10)] The expression for the noise covariance matrix at ℓ>52 is notationally confusing; the block structure and the role of the fiducial θ0 would benefit from a clearer derivation, including the assumption that the covariance is independent of θ.
[§III, Fig. 2 caption] The caption says 'obvious (anti-)correlations between parameters lead to the spikes at the boundaries of histogram.' This is a general statement that is not backed by the formal properties of rank statistics; please either justify it or rephrase, as it appears to contradict the standard interpretation of boundary spikes as miscalibration.
[Table I] The reported uncertainties are all Gaussian-symmetric at the precision shown, but the marginal posteriors in Fig. 5 may be asymmetric; please report asymmetric credible intervals where appropriate, as is common in cosmological parameter estimation.

Circularity Check

1 steps flagged · score 4.0 of 10

The inference pipeline is standard SBI and not circular, but the validation pass that accepts the w0 result imports its key 'anti-correlation spike' exception from the authors' own prior work.

self citation load bearing [Section III (validation text after Eq. 19) and Fig. 2 caption; Ref. [25]]
"there is an obvious left or right spike for H0 or w0, which results from the anti-correlation between them. ... The distribution of rank statistic and its deviation from U(0,1) can diagnose the incorrectness of estimated posterior [25, 35–37]: ... obvious (anti-)correlations between parameters lead to the spikes at the boundaries of histogram."

To declare the learned posterior 'credible' and thereby support the headline w0 = -0.90 ± 0.05, the paper must explain away the non-uniform rank statistic for w0 (and H0) as an anti-correlation artifact. The only specific citation attached to that exception is Ref. [25], Wang & Feng, which is by the same authors. The independent references [35-37] establish the general uniform-rank criterion for exact posteriors but do not establish that anti-correlation produces the boundary spikes. Thus the load-bearing validation premise is imported from the authors' own prior work rather than independently demonstrated.

full rationale

The core method is standard simulation-based inference: a wiCDM forward model built on CLASS simulates CMB, BAO, and SNIa data; SNLE is trained on 6x20000 simulations; and the learned posterior is evaluated on Planck 2018 + DESI DR2 + Pantheon+. This is not a circular derivation: the w_i constraints are generated by the forward model and data, not fitted to the claimed result. Validation via rank statistics, P-P plots, and predicted-vs-true comparisons checks internal consistency of the learned posterior, which is a legitimate but non-circular check. The single load-bearing self-citation is the interpretation of the H0/w0 rank-statistic spikes as harmless anti-correlation artifacts; that interpretation is attributed to Ref. [25] by the same authors. Since the headline claim of marginal dynamical dark energy depends on trusting the w0 marginal posterior, and the specific exception to uniform rank statistics is sourced from the authors' own prior work, the validation step is self-citation load-bearing. Nevertheless, the central constraint retains independent content from the forward simulations and external cosmological data, so the overall circularity is partial rather than complete.

Assumptions & free parameters 16 free parameters · 7 assumptions · 0 invented entities

The reconstruction rests on the piecewise-constant dark-energy model, the flat-universe assumption, the accuracy of CLASS, the fidelity of the simulated noise realizations, the convergence of the SNLE network, and the validity of the published data covariance matrices. The free parameters are the seven w_i, the absolute magnitude M, the six base ΛCDM parameters, the chosen prior bounds, and the fiducial cosmology used to build the CMB noise covariance. No new physical entities (particles, forces, dimensions) are introduced.

free parameters (16)

w0 = -0.90 ± 0.05
EOS in bin (0,0.4]; fitted from combined data; central to the marginal dynamical-DE claim.
w1 = -0.75 ± 0.37
EOS in bin (0.4,0.6]; consistent with -1 at 68% C.L.
w2 = -1.07 +0.82 -0.86
EOS in bin (0.6,0.8]; posterior broader than true according to rank statistics.
w3 = -1.90 +1.14 -1.04
EOS in bin (0.8,1.1].
w4 = -0.19 +1.27 -1.32
EOS in bin (1.1,1.4].
w5 = -6.16 +5.10 -12.36
EOS in bin (1.4,1.9]; unconstrained by data.
w6 = -38.66 +28.30 -38.18
EOS in bin (1.9,∞); unconstrained by data, posterior partly reflects very wide prior U(-100,0).
M = -19.41 ± 0.02
Absolute magnitude of SNIa, fitted nuisance parameter.
omega_b = 0.02221 ± 0.00014
Base ΛCDM parameter fitted alongside w_i.
omega_c = 0.1221 ± 0.0012
Base ΛCDM parameter.
H0 = 67.22 +0.56 -0.62 km/s/Mpc
Base ΛCDM parameter; anti-correlated with w0.
ln(10^10 As) = 3.059 ± 0.007
Base ΛCDM parameter.
ns = 0.9578 ± 0.0035
Base ΛCDM parameter.
tau = 0.0581 +0.0042 -0.0036
Base ΛCDM parameter.
Prior bounds for w_i = w0 U(-3,0), w1 U(-3,3), w2-w4 U(-10,3), w5 U(-50,3), w6 U(-100,0)
Chosen by hand; the very wide priors at high z directly shape the unconstrained w5/w6 posteriors.
Fiducial cosmology θ0 for CMB noise covariance = {0.02237,0.1200,1.04092,3.044,0.9649,0.0544}
Chosen from ΛCDM best fit; used to construct noise covariance in Eqs. (6)-(10); paper argues choice is model-independent if spectra fit Planck well.

assumptions (7)

domain assumption Spatial flatness (no curvature term in Friedmann equation)
Eq. (4) omits Ω_k; the analysis assumes a flat universe.
domain assumption The true w(z) is piecewise constant in the chosen 7 bins
Eq. (1) and the reconstruction method; if w varies within a bin, the reconstructed w_i is a biased average.
domain assumption CLASS accurately computes CMB spectra, distance ratios, and distances for w_iCDM
Sections II A-II C rely on CLASS as the exact forward model.
ad hoc to paper CMB noise realizations sampled from Wishart/multivariate normal with fiducial ΛCDM covariance correctly represent Planck 2018 measurement noise
Section II A, Eqs. (6)-(10); uses a ΛCDM fiducial cosmology and MontePython fake_planck_realistic noise.
domain assumption SNLE network trained for 6 rounds converges to the true likelihood of the forward simulator
Section III; the validity of the inference relies on the neural density estimator approximating q_W(x|θ).
domain assumption Pantheon+ and DESI DR2 covariance matrices (including systematics) are correct; Cholesky sampling from them is valid
Sections II B and II C.
domain assumption Uniform priors over the stated ranges represent ignorance; the constrained w0-w4 posteriors are not sensitive to the prior bounds
Eq. (5); the paper uses wide priors, but for the unconstrained w5/w6 the prior bounds affect the reported posterior.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Implicit Likelihood Inference and $z$-Binned Reconstruction of Dark Energy $w(z)$." pith.science (2026). https://pith.science/paper/4EIYDZGJ

@misc{pith2026260808007,
  author       = {Pith},
  title        = {Pith review of: Implicit Likelihood Inference and $z$-Binned Reconstruction of Dark Energy $w(z)$},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4EIYDZGJ}},
  note         = {Machine review of arXiv:2608.08007}
}

abstract

In this paper, to reconstruct the equation of state (EOS) of dark energy (DE) $w(z)$ with the redshift binning method, we first introduce a $w_i$CDM model with a piecewise-constant EOS in $7$ redshift bins. Then, we turn to the Learning the Universe Implicit Likelihood Inference (LtU-ILI) pipeline to perform a multi-round ILI of $w_i$ from the cosmological data combination, including $TT$, $TE$, $EE$ and lensing power spectra of Planck 2018, distance ratios of DESI DR2 and corrected apparent magnitudes of SNIa from Pantheon+ sample. More precisely, we build the Cosmic Microwave Background (CMB) power spectrum, Baryon Acoustic Oscillation (BAO) distance ratio and Type Ia Supernovae (SNIa) apparent magnitude simulators by $\mathtt{CLASS}$ and embed them into the LtU-ILI pipeline. And, using Sequential Neural Likelihood Estimation (SNLE), we sequentially train neural networks with $6$ rounds of total $6\times20000$ simulations to target a ``black box'' likelihood of our forward model $w_i$CDM. Finally, with the estimated posteriors of $w_i$, we find that except for the unconstrained $w_5$ and $w_6$ (the last two bins), our reconstruction of $w(z)$ marginally favors dynamical DE in the first bin and is consistent with the cosmological constant at $68\%$ C.L. in the other bins.

Figures

Figures reproduced from arXiv: 2608.08007 by the authors.

**Figure 2.** FIG. 2: Rank statistic of [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** FIG. 3: Marginal percentile coverage test for [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: FIG. 4: Comparison between predicted and true values of [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

**Figure 5.** Figure 5: FIG. 5: Constraints on parameters of [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

**Figure 6.** Figure 6: FIG. 6: Reconstruction of [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 7 canonical work pages

[1]

Observa- tional evidence from supernovae for an accelerating uni- verse and a cosmological constant,

A. G. Riesset al.[Supernova Search Team], “Observa- tional evidence from supernovae for an accelerating uni- verse and a cosmological constant,” Astron. J.116, 1009- 9 FIG. 5: Constraints on parameters ofw i from Planck 2018 [6], DESI DR2 [7] and Pantheon+ sample [28], where 68% interval of each parameter is displayed as the titles of diagonal subplots an...

arXiv 1998
[2]

According to the estimator of power spectra ˆC XX ′ ℓ (θ) = 1 2ℓ+ 1 ℓX m=−ℓ aX∗ ℓm aX ′ ℓm,(7) there obviously is a fundamental uncertainty (due to cos- mic variance) in knowledge about the power spectra. In 3 principle, the multipole coefficientsa X ℓm of CMB maps are assumed to follow a Gaussian distribution and can be sampled according to the covarianc...

work page 2018
[3]

As shown in Fig

The rank statistics is based on the conclusion that if the true posterior is exactly estimated, the rank statistic ofθ i should be distributed uniformly [25, 35–37]. As shown in Fig. 2, there is an obvious left or right spike forH 0 orw 0, which results from the anti-correlation between them. Except for spikes, the individual rank statistic of H0,w 0 orw ...

work page 2018
[4]

Exploring the expansion history of the uni- verse,

E. V. Linder, “Exploring the expansion history of the uni- verse,” Phys. Rev. Lett.90, 091301 (2003) [arXiv:astro- ph/0208512 [astro-ph]]

arXiv 2003
[5]

Measurements of Ω and Λ from 42 High Redshift Super- novae,

S. Perlmutteret al.[Supernova Cosmology Project], “Measurements of Ω and Λ from 42 High Redshift Super- novae,” Astrophys. J.517, 565-586 (1999) [arXiv:astro- ph/9812133 [astro-ph]]

arXiv 1999
[6]

Accelerating universes with scaling dark matter,

M. Chevallier and D. Polarski, “Accelerating universes with scaling dark matter,” Int. J. Mod. Phys. D10, 213- 224 (2001) [arXiv:gr-qc/0009008 [gr-qc]]

arXiv 2001
[7]

DESI DR2 results. II. Measurements of baryon acoustic oscillations and cos- mological constraints,

M. Abdul Karimet al.[DESI], “DESI DR2 results. II. Measurements of baryon acoustic oscillations and cos- mological constraints,” Phys. Rev. D112, no.8, 083515 (2025) [arXiv:2503.14738 [astro-ph.CO]]

arXiv 2025
[8]

A. G. Riess, W. Yuan, L. M. Macri, D. Scolnic, D. Brout, S. Casertano, D. O. Jones, Y. Murakami, L. Breuval and T. G. Brink,et al.“A Comprehensive Measurement of 10 FIG. 6: Reconstruction ofw(z) withw i, wherew 5 andw 6 were not constrained well so that they deviate from−1 a lot and don’t appear in plot, the shaded regions are 68% C.L. and the white lines...

arXiv 2022

Show all 41 references

[9]

Planck 2018 results. VI. Cosmological parameters,

N. Aghanimet al.[Planck], “Planck 2018 results. VI. Cosmological parameters,” Astron. Astrophys.641, A6 (2020) [erratum: Astron. Astrophys.652, C4 (2021)] [arXiv:1807.06209 [astro-ph.CO]]

2020 arXiv
[10]

Reconstruc- tion of dark energy and late-time cosmic expansion using the Weighted Function Regression method,

A. Gonz´ alez-Fuentes and A. G´ omez-Valent, “Reconstruc- tion of dark energy and late-time cosmic expansion using the Weighted Function Regression method,” JCAP12, 049 (2025) [arXiv:2506.11758 [astro-ph.CO]]

2025
[11]

Reconstruction of dark energy and expansion dynamics using Gaussian processes,

M. Seikel, C. Clarkson and M. Smith, “Reconstruction of dark energy and expansion dynamics using Gaussian processes,” JCAP06, 036 (2012) [arXiv:1204.2832 [astro- ph.CO]]

2012 arXiv
[12]

Reconstructing equation of state of dark energy with principal component analysis,

Z. E. Liu, H. F. Qin, J. Zhang, T. J. Zhang and H. R. Yu, “Reconstructing equation of state of dark energy with principal component analysis,” Phys. Dark Univ.26, 100379 (2019) [arXiv:1501.02971 [astro-ph.CO]]

2019 arXiv
[13]

Distance Pri- ors from Planck Final Release,

L. Chen, Q. G. Huang and K. Wang, “Distance Pri- ors from Planck Final Release,” JCAP02, 028 (2019) [arXiv:1808.05724 [astro-ph.CO]]

2019 arXiv
[14]

Reconstructing dark energy with model independent methods after DESI DR2,

J. X. Li and S. Wang, “Reconstructing dark energy with model independent methods after DESI DR2,” Eur. Phys. J. C85, no.11, 1308 (2025) [arXiv:2506.22953 [astro-ph.CO]]

2025
[15]

How the dark energy can reconcile Planck with local determination of the Hub- ble constant,

Q. G. Huang and K. Wang, “How the dark energy can reconcile Planck with local determination of the Hub- ble constant,” Eur. Phys. J. C76, no.9, 506 (2016) [arXiv:1606.05965 [astro-ph.CO]]

2016 arXiv
[16]

Reconstruc- tion of the Dark Energy equation of state from latest data: the impact of theoretical priors,

F. Gerardi, M. Martinelli and A. Silvestri, “Reconstruc- tion of the Dark Energy equation of state from latest data: the impact of theoretical priors,” JCAP07, 042 (2019) [arXiv:1902.09423 [astro-ph.CO]]

2019 arXiv
[17]

Dynamical dark energy in light of the latest observations,

G. B. Zhao, M. Raveri, L. Pogosian, Y. Wang, R. G. Crit- tenden, W. J. Handley, W. J. Percival, F. Beutler, J. Brinkmann and C. H. Chuang,et al.“Dynamical dark energy in light of the latest observations,” Nature Astron. 1, no.9, 627-632 (2017) [arXiv:1701.08165 [astro-ph.CO]]

2017 arXiv
[18]

Reconstruction of the Dark Energy Equation of State from the Latest Obser- vations,

J. P. Dai, Y. Yang and J. Q. Xia, “Reconstruction of the Dark Energy Equation of State from the Latest Obser- vations,” Astrophys. J.857, no.1, 9 (2018)

2018
[19]

The Cosmic Lin- ear Anisotropy Solving System (CLASS) II: Approxima- tion schemes,

D. Blas, J. Lesgourgues and T. Tram, “The Cosmic Lin- ear Anisotropy Solving System (CLASS) II: Approxima- tion schemes,” JCAP07, 034 (2011) [arXiv:1104.2933 [astro-ph.CO]]

2011 arXiv
[20]

Dynamical dark energy in light of the DESI DR2 baryonic acoustic oscillations mea- surements,

G. Guet al.[DESI], “Dynamical dark energy in light of the DESI DR2 baryonic acoustic oscillations mea- surements,” Nature Astron.9, no.12, 1879-1889 (2025) [erratum: Nature Astron.9, no.12, 1898-1898 (2025)] [arXiv:2504.06118 [astro-ph.CO]]

2025
[21]

The frontier of simulation-based inference,

K. Cranmer, J. Brehmer and G. Louppe, “The frontier of simulation-based inference,” Proc. Nat. Acad. Sci.117, no.48, 30055-30062 (2020) [arXiv:1911.01429 [stat.ML]]

2020 arXiv
[22]

Exploring cosmic origins with CORE: Cosmological parameters,

E. Di Valentinoet al.[CORE], “Exploring cosmic origins with CORE: Cosmological parameters,” JCAP04, 017 (2018) [arXiv:1612.00021 [astro-ph.CO]]

2018 arXiv
[23]

Efficient compu- tation of CMB anisotropies in closed FR W models,

A. Lewis, A. Challinor and A. Lasenby, “Efficient compu- tation of CMB anisotropies in closed FR W models,” As- trophys. J.538, 473-476 (2000) [arXiv:astro-ph/9911177 [astro-ph]]

2000 arXiv
[24]

Fast and credible likelihood-free cosmology with truncated marginal neural ratio estimation,

A. Cole, B. K. Miller, S. J. Witte, M. X. Cai, 11 M. W. Grootes, F. Nattino and C. Weniger, “Fast and credible likelihood-free cosmology with truncated marginal neural ratio estimation,” JCAP09, 004 (2022) [arXiv:2111.08030 [astro-ph.CO]]

2022 arXiv
[25]

Implicit likelihood inference of the neutrino mass hierarchy from cosmological data,

K. Wang and J. Y. Feng, “Implicit likelihood inference of the neutrino mass hierarchy from cosmological data,” Phys. Rev. D114, no.2, 023502 (2026) [arXiv:2512.17744 [astro-ph.CO]]

2026 arXiv
[26]

Conservative Constraints on Early Cosmology: an il- lustration of the Monte Python cosmological parameter inference code,

B. Audren, J. Lesgourgues, K. Benabed and S. Prunet, “Conservative Constraints on Early Cosmology: an il- lustration of the Monte Python cosmological parameter inference code,” JCAP02, 001 (2013) [arXiv:1210.7183 [astro-ph.CO]]

2013 arXiv
[27]

Likelihood methods for the combined analysis of CMB temperature and polari- sation power spectra,

W. J. Percival and M. L. Brown, “Likelihood methods for the combined analysis of CMB temperature and polari- sation power spectra,” Mon. Not. Roy. Astron. Soc.372, 1104-1116 (2006) [arXiv:astro-ph/0604547 [astro-ph]]

2006 arXiv
[28]

The Pantheon+ Analysis: Cosmolog- ical Constraints,

D. Brout, D. Scolnic, B. Popovic, A. G. Riess, J. Zuntz, R. Kessler, A. Carr, T. M. Davis, S. Hinton and D. Jones,et al.“The Pantheon+ Analysis: Cosmolog- ical Constraints,” Astrophys. J.938, no.2, 110 (2022) [arXiv:2202.04077 [astro-ph.CO]]

2022 arXiv
[29]

Planck 2018 results. VIII. Gravitational lensing,

N. Aghanimet al.[Planck], “Planck 2018 results. VIII. Gravitational lensing,” Astron. Astrophys.641, A8 (2020) [arXiv:1807.06210 [astro-ph.CO]]

2020 arXiv
[30]

https://pla.esac.esa.int/pla/#cosmology
[31]

sbi: A toolkit for simulation- based inference,

A. Tejero-Canteroet al., “sbi: A toolkit for simulation- based inference,” Journal of Open Source Software,5, no.52, 2505, (2020) [arXiv:2007.09114 [cs.LG]]

2020 arXiv
[32]

LtU-ILI: An All-in-One Frame- work for Implicit Inference in Astrophysics and Cos- mology,

M. Ho, D. J. Bartlett, N. Chartier, C. Cuesta-Lazaro, S. Ding, A. Lapel, P. Lemos, C. C. Lovell, T. L. Maki- nen and C. Modi,et al.“LtU-ILI: An All-in-One Frame- work for Implicit Inference in Astrophysics and Cos- mology,” Open J. Astrophys.7, 001c.120559 (2024) [arXiv:2402.0...

2024 arXiv
[33]

PyTorch: An Imperative Style, High- Performance Deep Learning Library,

A. Paszkeet al., “PyTorch: An Imperative Style, High- Performance Deep Learning Library,” [arXiv:1912.01703 [cs.LG]]

1912 arXiv
[34]

Masked Autoregressive Flow for Density Estima- tion,

G. Papamakarios, T, Pavlakou and I. Murray, “Masked Autoregressive Flow for Density Estima- tion,” [arXiv:1705.07057 [stat.ML]]

arXiv
[35]

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

J. Alsing, B. Wandelt and S. Feeney, “Massive op- timal data compression and density estimation for scalable, likelihood-free inference in cosmology,” Mon. Not. Roy. Astron. Soc.477, no.3, 2874-2885 (2018) [arXiv:1801.01497 [astro-ph.CO]]

2018 arXiv
[36]

Neural Density Estimation and Likelihood-free Inference,

G. Papamakarios, “Neural Density Estimation and Likelihood-free Inference,” [arXiv:1910.13233 [stat.ML]]

1910 arXiv
[37]

Accelerated Bayesian SED Modeling Using Amortized Neural Posterior Estimation,

C. Hahn and P. Melchior, “Accelerated Bayesian SED Modeling Using Amortized Neural Posterior Estimation,” Astrophys. J.938, no.1, 11 (2022) [arXiv:2203.07391 [astro-ph.GA]]

2022 arXiv
[38]

SimBIG: mock challenge for a forward modeling approach to galaxy clus- tering,

C. Hahn, M. Eickenberg, S. Ho, J. Hou, P. Lemos, E. Massara, C. Modi, A. Moradinezhad Dizgah, B. R. S. Blancard and M. M. Abidi, “SimBIG: mock challenge for a forward modeling approach to galaxy clus- tering,” JCAP04, 010 (2023) [arXiv:2211.00660 [astro- ph.CO]]

2023 arXiv
[39]

Validating Bayesian Inference Algorithms with Simulation-Based Calibration,

S. Talts, M. Betancourt, D. Simpson, A. Vehtari and A. Gelman, “Validating Bayesian Inference Algorithms with Simulation-Based Calibration,” [arXiv:1804.06788 [stat.ME]]

arXiv
[41]

emcee: The MCMC Hammer,

D. Foreman-Mackey, D. W. Hogg, D. Lang and J. Good- man, “emcee: The MCMC Hammer,” Publ. Astron. Soc. Pac.125, 306-312 (2013) [arXiv:1202.3665 [astro- ph.IM]]

2013 arXiv
[1701]

n1701   =L 1701   N1(0,1)

Then the noise realizations can be efficiently sampled from   n1 ... n1701   =L 1701   N1(0,1) ... N1701(0,1)   .(17) Finally, the SNIa apparent magnitude simulator com- binesw iCDM’s theoretical predictions ofm(z) at the same 1701 redshifts byCLASS[19] and the...

Pith tools

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

[1] [1]

Observa- tional evidence from supernovae for an accelerating uni- verse and a cosmological constant,

A. G. Riesset al.[Supernova Search Team], “Observa- tional evidence from supernovae for an accelerating uni- verse and a cosmological constant,” Astron. J.116, 1009- 9 FIG. 5: Constraints on parameters ofw i from Planck 2018 [6], DESI DR2 [7] and Pantheon+ sample [28], where 68% interval of each parameter is displayed as the titles of diagonal subplots an...

arXiv 1998

[2] [2]

According to the estimator of power spectra ˆC XX ′ ℓ (θ) = 1 2ℓ+ 1 ℓX m=−ℓ aX∗ ℓm aX ′ ℓm,(7) there obviously is a fundamental uncertainty (due to cos- mic variance) in knowledge about the power spectra. In 3 principle, the multipole coefficientsa X ℓm of CMB maps are assumed to follow a Gaussian distribution and can be sampled according to the covarianc...

work page 2018

[3] [3]

As shown in Fig

The rank statistics is based on the conclusion that if the true posterior is exactly estimated, the rank statistic ofθ i should be distributed uniformly [25, 35–37]. As shown in Fig. 2, there is an obvious left or right spike forH 0 orw 0, which results from the anti-correlation between them. Except for spikes, the individual rank statistic of H0,w 0 orw ...

work page 2018

[4] [4]

Exploring the expansion history of the uni- verse,

E. V. Linder, “Exploring the expansion history of the uni- verse,” Phys. Rev. Lett.90, 091301 (2003) [arXiv:astro- ph/0208512 [astro-ph]]

arXiv 2003

[5] [5]

Measurements of Ω and Λ from 42 High Redshift Super- novae,

S. Perlmutteret al.[Supernova Cosmology Project], “Measurements of Ω and Λ from 42 High Redshift Super- novae,” Astrophys. J.517, 565-586 (1999) [arXiv:astro- ph/9812133 [astro-ph]]

arXiv 1999

[6] [6]

Accelerating universes with scaling dark matter,

M. Chevallier and D. Polarski, “Accelerating universes with scaling dark matter,” Int. J. Mod. Phys. D10, 213- 224 (2001) [arXiv:gr-qc/0009008 [gr-qc]]

arXiv 2001

[7] [7]

DESI DR2 results. II. Measurements of baryon acoustic oscillations and cos- mological constraints,

M. Abdul Karimet al.[DESI], “DESI DR2 results. II. Measurements of baryon acoustic oscillations and cos- mological constraints,” Phys. Rev. D112, no.8, 083515 (2025) [arXiv:2503.14738 [astro-ph.CO]]

arXiv 2025

[8] [8]

A. G. Riess, W. Yuan, L. M. Macri, D. Scolnic, D. Brout, S. Casertano, D. O. Jones, Y. Murakami, L. Breuval and T. G. Brink,et al.“A Comprehensive Measurement of 10 FIG. 6: Reconstruction ofw(z) withw i, wherew 5 andw 6 were not constrained well so that they deviate from−1 a lot and don’t appear in plot, the shaded regions are 68% C.L. and the white lines...

arXiv 2022