Pith. sign in

REVIEW 3 major objections 4 minor 79 references

This paper claims that combining a time-evolving dark energy equation of state (the CPL form) with a free summed neutrino mass can reduce the Hubble tension to roughly 1–2σ relative to both early-universe CMB and local distance-ladder measu

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 13:03 UTC pith:EHQYBPMO

load-bearing objection The central 'balanced reconciliation' claim doesn't survive contact with the paper's own equations: the CMB distance-prior computation uses a radiation-free H(z) extrapolated to z*≈1090, a load-bearing internal inconsistency that invalidates the reported H0 and tension numbers. the 3 major comments →

arxiv 2601.00495 v2 pith:EHQYBPMO submitted 2026-01-01 astro-ph.CO gr-qc

Late-Time Alleviation of the Hubble Tension in CPL Cosmology with Massive Neutrinos via Bayesian Physics-Informed Neural Networks

Muhammad Yarahmadi This is my paper
classification astro-ph.CO gr-qc
keywords Hubble tensionCPL parametrizationmassive neutrinosBayesian physics-informed neural networksdark energy equation of stateH0 inferencebaryon acoustic oscillationscosmic chronometers
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper tries to establish that a late-time-only modification of the standard cosmology—dark energy with a time-evolving equation of state plus a free summed neutrino mass—can bring early- and late-universe determinations of the Hubble constant into much better agreement than ΛCDM does. Using a Bayesian physics-informed neural network that hard-wires the Friedmann equation into the fit, the authors combine CMB distance priors with cosmic-chronometer, baryon-acoustic-oscillation, and supernova data. They find that the most flexible CPL+Σmν scenario yields H0 around 70–71.6 km s−1 Mpc−1, with tension typically at the 1–2σ level relative to both the early-universe CMB value and the local distance ladder, and tight upper bounds on the neutrino mass sum (<0.16–0.28 eV). The paper further claims that the neural-network inference reproduces a full MCMC analysis within 1–2σ at a fraction of the computational cost.

Core claim

In the paper's own terms, the central discovery is that the CPL+Σmν model—dark energy with equation of state w(z)=w0+wa z/(1+z) plus a free summed neutrino mass treated as an effective late-time matter component—reconciles early- and late-universe determinations of H0 better than ΛCDM does. The evolving equation of state systematically moves H0 upward relative to ΛCDM, especially in combinations containing supernova data, while the free neutrino mass acts as a stabilizing ingredient that keeps the global fit coherent and yields conservative upper bounds Σmν<0.16–0.28 eV at 1σ. For the full combination of CMB priors, cosmic chronometers, BAO, and supernovae, the inferred H0 is 70.43±1.20 km s

What carries the argument

The central object is the Bayesian physics-informed neural network (BPINN) representation of the dimensionless Hubble rate E(z)=H(z)/H0. The network is trained with a soft physics loss that enforces E²(z)=Ωm0(1+z)³ + (1−Ωm0)(1+z)^{3(1+w0+wa)} exp(−3wa z/(1+z)), the CPL Friedmann equation with neutrinos folded into the effective matter density Ωm0. Dropout layers provide approximate Bayesian uncertainty quantification, and the same network output is used to compute comoving distances and the sound horizon entering the CMB shift parameter R and acoustic scale lA. This object carries the argument because it lets the authors vary w0, wa, Σmν, and H0 jointly while keeping the expansion history co

Load-bearing premise

Every fit includes early-universe CMB distance priors, but the network's background equation ignores radiation and treats massive neutrinos as ordinary matter at all redshifts—including at z≈1090 where those approximations fail—so if the extrapolated sound horizon and distance are biased, all reported H0 values and tension levels shift.

What would settle it

Take the best-fit parameters from any CPL+Σmν dataset combination and compute the CMB shift parameter R and acoustic scale lA with a full radiation-plus-neutrino Boltzmann calculation rather than the PINN's matter-only Friedmann model; if the predicted (R, lA) move away from the measured CMB values by more than the prior covariance, or if re-running the same likelihood changes H0 by more than about 1σ, the claimed 1–2σ reconciliation is an artifact of the simplified early-universe model.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If the claim holds, late-time-only physics can reduce a >5σ discrepancy to roughly 1–2σ, weakening the default assumption that the Hubble tension requires new physics before recombination.
  • The preferred mildly phantom equation of state (w0≈−1.05 to −1.12, wa≈0.01–0.03) implies dark energy crosses the phantom divide at low redshift, which no single minimally coupled scalar field can do; the model must be read as an effective description.
  • The neutrino mass sum is not detected, but the consistent upper bounds around 0.16–0.28 eV are tight enough to be tested by upcoming CMB and large-scale-structure surveys.
  • Because the same BPINN reproduces MCMC posteriors within 1–2σ, the machine-learning method itself is unlikely to be the source of the reported alleviation.
  • The persistence of ~2.5σ tension with the early-universe CMB value in the most constraining combinations means late-time extensions alone may not be the full story.

Where Pith is reading between the lines

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

  • I infer that the reported tension numbers hinge on treating neutrinos as matter-like at all redshifts: a Boltzmann-level recalculation of the sound horizon could shift H0 by more than the quoted error bars, and would settle whether the 1–2σ reconciliation is physical or a modeling artifact.
  • I infer that the phantom-crossing preference could be absorbed by interacting dark energy or modified gravity; adding growth-rate data (fσ8, S8) would test whether the same model fixes or worsens the growth tension.
  • I infer that the method's computational advantage makes it attractive for scanning large model spaces, but the simplified early-universe model means it should be paired with a full Boltzmann recalculation before claiming precision cosmological constraints.
  • I infer that the effective late-time Σmν bounds should not be read as particle-physics neutrino mass measurements; a full treatment would broaden or shift them.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper presents Bayesian Physics-Informed Neural Network (BPINN) analyses of ΛCDM, CPL, and CPL+Σmν cosmologies, combining cosmic chronometers, DESI DR2 BAO, Pantheon+ SNe, and Planck 2018 CMB distance priors. The central claim is that the most flexible model, CPL with free Σmν, shifts H0 to ~70–71.6 km/s/Mpc and reduces the Hubble tension to ~1–2σ relative to both Planck and SH0ES. The authors also compare their BPINN results to a full-dataset MCMC analysis and report broad agreement. The underlying methodology enforces the Friedmann equation as a soft PINN constraint and propagates uncertainties through stochastic dropout.

Significance. If the quantitative results were reliable, the paper would provide evidence that late-time dark-energy dynamics plus neutrino mass can substantially relieve the Hubble tension without early-time new physics. The intended contribution—a Bayesian PINN framework for background cosmology with uncertainty propagation—is potentially useful. However, the central quantitative claim is not supported by the model as written: the CMB distance-prior computation described in Sec. V.B is inconsistent with the Friedmann constraint in Eq. (13). Because every dataset combination in Tables I–III includes the CMB priors, this error is load-bearing. The paper does not provide code or a reproducible pipeline, which limits independent verification.

major comments (3)
  1. [Sec. V.B and Eq. (13)] The paper states that at every sampling step the PINN-predicted H(z) is used to compute D_C(z*) and r_s(z*) for R and l_A. However, Eq. (13) is E² = Ωm(1+z)³ + ΩDE f_CPL(z), with radiation neglected and massive neutrinos absorbed into Ωm as matter at all redshifts. At z*≈1090 radiation contributes ~25–30% of H², and neutrinos with Σmν≈0.2 eV are relativistic at that epoch. A literal evaluation of Eq. (13) at the Table II full-combination best fit (Ωm≈0.308, H0≈70.4) gives R≈1.94, versus the Planck prior R=1.7502 used in Eq. (39)—roughly 30σ away. Since all likelihoods include the CMB distance priors, the reported H0 values and tension statistics are not supported by the model described.
  2. [Sec. II.B.a and Eq. (34)] The treatment of the CMB prior is self-contradictory. The text says the Planck distance priors are adopted 'assuming standard early-time physics and a fixed sound horizon,' but Sec. V.B says r_s(z*) is recomputed from the PINN-predicted H(z) via Eq. (34). With the PINN H(z) from Eq. (13), the early-universe sound horizon integral is matter-dominated and gives r_s ∝ a*^{1/2}, not the standard radiation-dominated result. One cannot simultaneously fix r_s to standard physics and recompute it from Eq. (13). This contradiction directly affects the acoustic scale l_A and hence the CMB term in Eq. (40).
  3. [Table IV] The reported CMB χ² values (0.01–0.34) are incompatible with the literal procedure in Sec. V.B. Given the R mismatch estimated above, Eq. (38) alone would yield χ²_CMB of order thousands before including l_A and ωb. Table IV therefore cannot describe the model defined by Eq. (13). Either the CMB priors were evaluated with an unstated standard early-universe calculation, or the tables are internally inconsistent. Either way, the central quantitative conclusions—especially the claimed 1–2σ tension alleviations in Table II—lack support.
minor comments (4)
  1. [Eq. (11)] The priors list Ων ∈ [0,0.5], but the sampled parameter is Σmν. Please clarify the conversion in Eq. (6) and whether h is held fixed when imposing this prior.
  2. [Table V] Table V lists only three free parameters for CPL (Ωm, H0, ΩDE), but the CPL model used elsewhere has four (Ωm, H0, w0, wa). The 'simplified CPL' label is not explained and the χ²/AIC/BIC comparison is therefore ambiguous.
  3. [References] Several references are incomplete or duplicate: e.g., Ref. [29] and [33] are the same paper; Ref. [57] gives only 'arXiv:1907.12875' without author/title details; Refs. [38–40] are formatted inconsistently. Please unify the bibliography.
  4. [Figures] Many corner-plot axis labels and legends are too small to read (e.g., Figs. 1, 3, 5, 7). The legends in Figs. 3, 5, and 7 do not clearly identify all dataset combinations.

Circularity Check

2 steps flagged

CMB distance-prior 'predictions' reduce to fixed Planck inputs: Eq. (13) cannot produce r_s(z*), and the reported R/l_A agreement is input-matching.

specific steps
  1. fitted input called prediction [Section V.B, paragraph 1; Eq. (39); footnote II.B.a]
    "At each step of the Bayesian sampling, the PINN-predicted Hubble function H(z) is used to compute the comoving distance D_C(z*) and the sound horizon r_s(z*), from which the CMB shift parameter R and the acoustic angular scale ℓ_A are derived. These predicted quantities are then compared with their Planck observational values ... The Planck 2018 CMB distance priors are therefore adopted assuming standard early-time physics and a fixed sound horizon."

    The Planck values R=1.7502 and l_A=301.471 are entered as the observed vector in Eq. (39), i.e. as inputs to the likelihood. The paper simultaneously claims that r_s(z*) is computed from the PINN-predicted H(z), but the footnote states that the CMB distance priors are adopted with a fixed sound horizon from standard early-time physics. Thus the 'predicted' R and l_A are not derived from the claimed CPL+Σmν model; they are effectively the same Planck inputs used in the χ²_CMB term. The reported χ²_CMB≈0.03 and the resulting 1–2σ 'reconciliation' with Planck are therefore input-matching residuals, not independent model predictions.

  2. other [Eq. (13) and Eq. (34); Section V.B; footnote II.B.a]
    "E^2(z) = Ωm,0(1+z)^3 + ΩDE,0 fCPL(z) ... Radiation is neglected in the above equation, as its contribution to the expansion rate is negligible over the redshift range probed by the BAO, cosmic chronometer, and supernova data sets considered in this work. ... r_s(z*) = ∫_0^{a*} c_s(a)/(a^2 H(a)) da."

    The paper says the PINN H(z) is used at each sampling step to compute r_s(z*) at z*≈1090. But Eq. (13) deliberately omits radiation and treats massive neutrinos as a matter-like component at all redshifts, so the extrapolated H(z) is not the physical expansion rate in the pre-recombination era that dominates the r_s integral. To obtain χ²_CMB values of order 0.01–0.03 with Planck's R and l_A, the computation must fall back on the fixed standard-early-time sound horizon admitted in footnote II.B.a. Therefore the early-Universe 'constraint' is not a first-principles prediction of the model as stated; the agreement with Planck is effectively inserted through the fixed input, making the central 'balanced reconciliation' partly circular by construction.

full rationale

Most of the paper is a standard Bayesian posterior analysis, and the many self-citations are not load-bearing for the main derivation. The central H0/tension numbers, however, depend on a claimed first-principles computation of the CMB distance priors from the PINN expansion history that is not actually possible with the stated model. Section V.B explicitly says the PINN-predicted H(z) is used to compute D_C(z*) and r_s(z*), while Eq. (13) omits radiation and treats massive neutrinos as matter at all redshifts; the sound-horizon integral at z*≈1090 is dominated by precisely the physics that Eq. (13) excludes. The paper's own footnote II.B.a concedes that the Planck 2018 CMB distance priors are adopted 'assuming standard early-time physics and a fixed sound horizon.' Thus the R and l_A entering χ²_CMB are effectively the Planck input values rather than predictions of the CPL+Σmν model. Similarly, the BAO sound-horizon scale is said to 'vary implicitly through its dependence on early-Universe physics' without any early-universe model in the PINN, so the Planck-anchored calibration is partly carried into the BAO term. As a result, the reported reduction of the Hubble tension relative to Planck is substantially constructed from the input priors rather than derived from the late-time model alone. The SH0ES comparison and the late-time data remain external, so the circularity is partial rather than total. Score 6 reflects this partial, load-bearing reduction of the central 'reconciliation' claim to the input CMB distance priors, alongside an internal model inconsistency that makes the literal computation impossible.

Axiom & Free-Parameter Ledger

6 free parameters · 6 axioms · 0 invented entities

The central claim rests on five fitted cosmological parameters plus a hand-chosen physics-loss weight, and on the ad hoc assumption that a late-time-only background model (no radiation, matter-like neutrinos) can be extrapolated to z*≈1090 to compute CMB distance priors. No new particles or forces are introduced.

free parameters (6)
  • H0 = 68.45–71.86 km/s/Mpc (depending on model and dataset)
    Central fitted parameter; values shift with model and data combination.
  • Ωm = 0.29–0.34
    Present-day matter density, fitted from distance and expansion data.
  • w0 = −1.02 to −1.16
    Present-day dark-energy equation-of-state parameter, fitted.
  • wa = 0.01–0.03
    Dark-energy equation-of-state time variation, fitted.
  • Σmν = upper limits 0.16–0.28 eV
    Effective late-time neutrino mass parameter, fitted; only enters background as matter-like contribution.
  • λ_phys = not specified
    Weight of the physics-informed loss in Eq. (18); a hand-chosen hyperparameter that controls how strongly the Friedmann constraint is enforced.
axioms (6)
  • standard math Spatially flat FLRW background
    Eq. (1)–(2) assume flatness; standard and observationally motivated, but not derived in this paper.
  • domain assumption Dark energy is separately conserved
    Eq. (4) assumes no interaction between dark energy and other components.
  • domain assumption Massive neutrinos treated as pressureless matter at late times
    Section II.B states neutrinos are effectively non-relativistic at z≲2; this is an approximation that fails at higher z.
  • ad hoc to paper Radiation negligible in the Friedmann constraint
    Eq. (13) neglects radiation; valid for z≲2 but used later for CMB priors at z*≈1090.
  • ad hoc to paper Standard early-time physics and fixed sound horizon for CMB distance priors
    Explicit limitation note in Section II.B: early-universe neutrino perturbations and sound horizon are not modeled; Planck priors assume standard early-time physics.
  • standard math Chen et al. (2019) CMB distance priors and covariance matrix are applicable
    Section IV.D adopts published compressed CMB distance priors; validity in CPL+Σmν models with effective late-time neutrino treatment is assumed, not demonstrated.

pith-pipeline@v1.3.0-alltime-deepseek · 25309 in / 17691 out tokens · 173223 ms · 2026-08-03T13:03:54.105703+00:00 · methodology

0 comments
read the original abstract

We present a comprehensive Bayesian analysis of the Hubble constant within the framework of Physics-Informed Neural Networks (PINNs), focusing on the standard $\Lambda$CDM model and its dynamical dark energy extensions described by the Chevallier-Polarski-Linder (CPL) parametrization, both with and without massive neutrinos. By embedding the cosmological background equations directly into a Bayesian PINN architecture, we reconstruct the Hubble expansion history $H(z)$ in a data-driven yet physically consistent manner, while rigorously propagating epistemic uncertainties. Our analysis combines late-time observational probes, including Cosmic Chronometers, Baryon Acoustic Oscillations (BAO DESI DR2), and the Pantheon supernova sample, and quantifies the resulting tension in the inferred Hubble constant with respect to Planck 2018 Cosmic Microwave Background constraints and the SH0ES (R22) local distance ladder measurement. Within $\Lambda$CDM, we find that data combinations involving BAO tend to favor lower values of $H_0$, alleviating the tension with Planck at the expense of increased disagreement with SH0ES. Allowing for a time-evolving dark energy equation of state in the CPL framework systematically shifts the posterior of $H_0$ toward higher values, leading to a notable reduction of the SH0ES tension, particularly for combinations including supernova data. The most flexible scenario, CPL with a free total neutrino mass $\Sigma m_\nu$, yields a balanced reconciliation between early- and late-Universe determinations of $H_0$, with tension levels typically reduced to the $\sim1$-$2\sigma$ range relative to both Planck and SH0ES. Our results highlight the nontrivial interplay between dark energy dynamics and neutrino mass in addressing the Hubble tension and demonstrate the efficacy of Bayesian PINNs as a robust and versatile tool for precision cosmology beyond the standard paradigm.

Figures

Figures reproduced from arXiv: 2601.00495 by Muhammad Yarahmadi.

Figure 1
Figure 1. Figure 1: Posterior distributions of the CPL dark energy parameters ( [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Comparison the Hubble constant H0 derived from the Bayesian Physics-Informed Neural Network (BPINN) analysis using various low-redshift datasets with Planck 2018 and R22. Physical Interpretation of the CPL Dark Energy Constraints The CPL parametrization offers a flexible phenomenological description to quantify deviations from a cosmological constant and to probe possible late–time dynamics of dark energy.… view at source ↗
Figure 3
Figure 3. Figure 3: Corner plot of the CPL+Σmν posterior distributions obtained from the Bayesian Physics-Informed Neural Network (BPINN) analysis. The filled contours correspond to the 68% and 95% confidence levels, showing the correlations between H0, Ωm, w0, wa, and the sum of neutrino masses Σmν. The posterior estimates of the Hubble constant H0 obtained from the BPINN analysis incorporating massive neutrinos are presente… view at source ↗
Figure 4
Figure 4. Figure 4: Comparison the Hubble constant H0 derived from the Bayesian Physics-Informed Neural Network (BPINN) analysis using various low-redshift datasets with Planck 2018 and R22. VI. BAYESIAN PHYSICS-INFORMED NEURAL NETWORK CONSTRAINTS ON THE ΛCDM MODEL In this section, we present the cosmological constraints obtained for the standard ΛCDM model using a Bayesian Physics-Informed Neural Network (BPINN) framework. T… view at source ↗
Figure 5
Figure 5. Figure 5: Corner plot of the ΛCDM posterior distributions obtained from the Bayesian Physics-Informed Neural Network [PITH_FULL_IMAGE:figures/full_fig_p021_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison the Hubble constant H0 derived from the Bayesian Physics-Informed Neural Network (BPINN) analysis using various low-redshift datasets with Planck 2018 and R22. Table III: Summary of ΛCDM parameter constraints and the corresponding Hubble tension levels with respect to Planck 2018 and SH0ES (R22). All uncertainties are quoted at the 1σ confidence level. Dataset H0 [km s−1 Mpc−1 ] Ωm w0 TPlanck [σ… view at source ↗
Figure 7
Figure 7. Figure 7: Corner plot of the CPL+Σmν posterior distributions obtained from the MCMC analysis. The filled contours correspond to the 68% and 95% confidence levels, showing the correlations between H0, Ωm, w0, wa, and the sum of neutrino masses Σmν. The results presented in Tables VI and II play a central role in assessing the consistency between different inference frameworks and the constraining power of the employe… view at source ↗

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

79 extracted references · 3 canonical work pages

  1. [1]

    Observational Evidence from Supernovae for an Accelerating Universe and a Cosmological Constant,

    A. G. Riesset al., “Observational Evidence from Supernovae for an Accelerating Universe and a Cosmological Constant,” Astron. J.116, 1009 (1998)

    1998

  2. [2]

    Measurements of Ω and Λ from 42 High-Redshift Supernovae,

    S. Perlmutteret al., “Measurements of Ω and Λ from 42 High-Redshift Supernovae,”Astrophys. J.517, 565 (1999)

    1999

  3. [3]

    The Cosmological Constant Problem,

    S. Weinberg, “The Cosmological Constant Problem,”Rev. Mod. Phys.61, 1 (1989)

    1989

  4. [4]

    The Cosmological Constant and Dark Energy,

    P. J. E. Peebles and B. Ratra, “The Cosmological Constant and Dark Energy,”Rev. Mod. Phys.75, 559 (2003)

    2003

  5. [5]

    Planck 2018 results. VI. Cosmological parameters,

    Planck Collaboration, “Planck 2018 results. VI. Cosmological parameters,”Astron. Astrophys.641, A6 (2020)

    2018

  6. [6]

    Dynamics of Dark Energy,

    E. J. Copeland, M. Sami and S. Tsujikawa, “Dynamics of Dark Energy,”Int. J. Mod. Phys. D15, 1753 (2006)

    2006

  7. [7]

    Modified Gravity and Cosmology,

    T. Clifton, P. G. Ferreira, A. Padilla and C. Skordis, “Modified Gravity and Cosmology,”Phys. Rept.513, 1 (2012)

    2012

  8. [8]

    Accelerating Universes with Scaling Dark Matter,

    M. Chevallier and D. Polarski, “Accelerating Universes with Scaling Dark Matter,”Int. J. Mod. Phys. D10, 213 (2001)

    2001

  9. [9]

    Exploring the Expansion History of the Universe,

    E. V. Linder, “Exploring the Expansion History of the Universe,”Phys. Rev. Lett.90, 091301 (2003)

    2003

  10. [10]

    Exploring the Expanding Universe and Dark Energy Using the Statefinder Diagnostic,

    U. Alam, V. Sahni, T. D. Saini and A. A. Starobinsky, “Exploring the Expanding Universe and Dark Energy Using the Statefinder Diagnostic,”Mon. Not. Roy. Astron. Soc.344, 1057 (2003)

    2003

  11. [11]

    Large Magellanic Cloud Cepheid Standards Provide a 1% Foundation for the Determination of the Hubble Constant,

    A. G. Riesset al., “Large Magellanic Cloud Cepheid Standards Provide a 1% Foundation for the Determination of the Hubble Constant,”Astrophys. J.876, 85 (2019)

    2019

  12. [12]

    Tensions between the Early and the Late Universe,

    L. Verde, T. Treu and A. G. Riess, “Tensions between the Early and the Late Universe,”Nature Astronomy3, 891 (2019)

    2019

  13. [13]

    In the Realm of the Hubble Tension—A Review of Solutions,

    E. Di Valentinoet al., “In the Realm of the Hubble Tension—A Review of Solutions,”Class. Quant. Grav.38, 153001 (2021)

    2021

  14. [14]

    Lower Growth Rate from Recent Redshift Space Distortion Measurements than Expected from Planck,

    E. Macaulay, I. K. Wehus and H. K. Eriksen, “Lower Growth Rate from Recent Redshift Space Distortion Measurements than Expected from Planck,”Phys. Rev. Lett.111, 161301 (2013)

    2013

  15. [15]

    KiDS-450: Cosmological Parameter Constraints from Tomographic Weak Gravitational Lensing,

    H. Hildebrandtet al., “KiDS-450: Cosmological Parameter Constraints from Tomographic Weak Gravitational Lensing,” Mon. Not. Roy. Astron. Soc.465, 1454 (2017)

    2017

  16. [16]

    Tension and Constraints on Modified Gravity Parametrizations ofGeff (z) from Growth Rate and Planck Data,

    S. Nesseris, G. Pantazis and L. Perivolaropoulos, “Tension and Constraints on Modified Gravity Parametrizations ofGeff (z) from Growth Rate and Planck Data,”Phys. Rev. D96, 023542 (2017)

    2017

  17. [17]

    Massive neutrinos and cosmology,

    J. Lesgourgues and S. Pastor, “Massive neutrinos and cosmology,”Phys. Rept.429, 307 (2006)

    2006

  18. [18]

    Neutrino mass from cosmology,

    J. Lesgourgues and S. Pastor, “Neutrino mass from cosmology,”Adv. High Energy Phys.2012, 608515 (2012)

    2012

  19. [19]

    Unveilingνsecrets with cosmological data: Neutrino masses and mass hierarchy,

    S. Vagnozziet al., “Unveilingνsecrets with cosmological data: Neutrino masses and mass hierarchy,”Phys. Rev. D98, 083501 (2018)

    2018

  20. [20]

    Beyond six parameters: extending ΛCDM,

    E. Di Valentino, A. Melchiorri and J. Silk, “Beyond six parameters: extending ΛCDM,”Phys. Rev. D95, 043502 (2017). 27

    2017

  21. [21]

    A CMB-based analysis of neutrino masses,

    M. Escudero and S. J. Witte, “A CMB-based analysis of neutrino masses,”Eur. Phys. J. C80, 294 (2020)

    2020

  22. [22]

    Physics–Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Differential Equations,

    M. Raissi, P. Perdikaris and G. E. Karniadakis, “Physics–Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Differential Equations,”J. Comput. Phys.378, 686 (2019)

    2019

  23. [23]

    B–PINNs: Bayesian Physics–Informed Neural Networks for Forward and Inverse PDE Problems with Uncertainty Quantification,

    L. Yang, X. Meng and G. E. Karniadakis, “B–PINNs: Bayesian Physics–Informed Neural Networks for Forward and Inverse PDE Problems with Uncertainty Quantification,”J. Comput. Phys.425, 109913 (2021)

    2021

  24. [24]

    The European Physical Journal C 83 (910 (2023))

    M Yarahmadi, A Salehi. The European Physical Journal C 83 (910 (2023))

    2023

  25. [25]

    Communications in Theoretical Physics 75 - 055401(2023)

    M Yarahmadi, A Salehi. Communications in Theoretical Physics 75 - 055401(2023)

    2023

  26. [26]

    Scientific Machine Learning through Physics–Informed Neural Networks: Where we are and What’s next,

    S. Cuomoet al., “Scientific Machine Learning through Physics–Informed Neural Networks: Where we are and What’s next,”J. Sci. Comput.92, 88 (2022)

    2022

  27. [27]

    Bayesian neural networks in cosmology: reconstruction and model comparison,

    S. Sharma, M. Kostrzewa–Rutkowska and J. Bayley, “Bayesian neural networks in cosmology: reconstruction and model comparison,”Mon. Not. Roy. Astron. Soc.526, 3526 (2023)

    2023

  28. [28]

    Dropout as a Bayesian Approximation,

    Y. Gal and Z. Ghahramani, “Dropout as a Bayesian Approximation,”ICML(2016)

    2016

  29. [30]

    Surrogate modeling for fluid flows,

    L. Sun, H. Gao, S. Pan and J. X. Wang, “Surrogate modeling for fluid flows,”Comput. Methods Appl. Mech. Eng.361, 112732 (2020)

    2020

  30. [31]

    The Pantheon+ Analysis: The Full Dataset and Cosmological Constraints,

    D. Scolnicet al., “The Pantheon+ Analysis: The Full Dataset and Cosmological Constraints,”Astrophys. J.938, 113 (2022)

    2022

  31. [32]

    Measuring cosmic bulk flow with Pantheon catalogue in perturbed f(R) gravity,

    M. Yarahmadi, A. Salehi, and H. Farajollahi, “Measuring cosmic bulk flow with Pantheon catalogue in perturbed f(R) gravity,”Mon. Not. R. Astron. Soc.527(4), 11840–11854 (2024)

    2024

  32. [33]

    A Comparative Bayesian PINN–MCMC Analysis of Barrow–Tsallis Holographic Dark Energy with Neutrinos: Toward Resolving the Hubble Tension,

    M. Yarahmadi and A. Salehi, “A Comparative Bayesian PINN–MCMC Analysis of Barrow–Tsallis Holographic Dark Energy with Neutrinos: Toward Resolving the Hubble Tension,”J. High Energy Astrophys., 100498 (2025)

    2025

  33. [35]

    Cosmic Bulk Flow Analysis in Modified Gravity Theories: and Perturbed Models with Neutrino Coupling,

    M. Yarahmadi and A. Salehi, “Cosmic Bulk Flow Analysis in Modified Gravity Theories: and Perturbed Models with Neutrino Coupling,”Astron. J., 10.3847/1538-3881/adaf16 (2025)

    work page doi:10.3847/1538-3881/adaf16 2025

  34. [36]

    A 2.4% Determination of the Local Value of the Hubble Constant,

    A. G. Riesset al., “A 2.4% Determination of the Local Value of the Hubble Constant,”Astrophys. J.826, 56 (2016)

    2016

  35. [37]

    A Comprehensive Measurement of the Local Value of the Hubble Constant with 1 km s −1 Mpc−1 Uncertainty from the Hubble Space Telescope and the SH0ES Team,

    A. G. Riesset al., “A Comprehensive Measurement of the Local Value of the Hubble Constant with 1 km s −1 Mpc−1 Uncertainty from the Hubble Space Telescope and the SH0ES Team,”Astrophys. J. Lett.934, L7 (2022)

    2022

  36. [38]

    M. G. Dainotti et al., Astrophys. J. 912, 150 (2021), arXiv:2103.02117 [astro-ph.CO]

    Pith/arXiv arXiv 2021

  37. [39]

    M. G. Dainotti et al., Galaxies 10, 24 (2022), arXiv:2201.09848 [astro-ph.CO]

    Pith/arXiv arXiv 2022

  38. [40]

    M. G. Dainotti et al., J. High Energy Astrophys. 48, 100405 (2025), arXiv:2501.11772 [astro-ph.CO]

    Pith/arXiv arXiv 2025

  39. [41]

    et al.,2020, Astrophys

    Moresco, M. et al.,2020, Astrophys. J. 898, 82 , arXiv:2003.07362 [astro-ph.GA]

    Pith/arXiv arXiv 2020

  40. [42]

    et al.,2012, J

    Moresco, M. et al.,2012, J. Cosmol. Astropart. Phys. 8, 006 , arXiv:1201.3609 [astro-ph.CO]

    Pith/arXiv arXiv 2012

  41. [43]

    Moresco,2015, M., Mon. Not. R. Astron. Soc. Lett. 450, L16 , arXiv:1503.01116 [astro-ph.CO]

    Pith/arXiv arXiv 2015

  42. [44]

    et al.,2016, J

    Moresco, M. et al.,2016, J. Cosmol. Astropart. Phys. 5, 014 , arXiv:1601.01701 [astro-ph.CO]

    Pith/arXiv arXiv 2016

  43. [45]

    Coupled non-canonical scalar field to neutrinos could alleviate the Hubble tension and cross the phantom barrier,

    M. Yarahmadi, A. Salehi, and A. Tohidi, “Coupled non-canonical scalar field to neutrinos could alleviate the Hubble tension and cross the phantom barrier,”Commun. Theor. Phys.(2025)

    2025

  44. [46]

    Neutrino Interactions with perturbed Rastall Gravity: A Novel Approach to Reducing the Hubble Ten- sion,

    M. Yarahmadi, “Neutrino Interactions with perturbed Rastall Gravity: A Novel Approach to Reducing the Hubble Ten- sion,”Int. J. Mod. Phys. A, 10.1142/S0217751X25500939 (2025)

    work page doi:10.1142/s0217751x25500939 2025

  45. [47]

    Generalized Logarithmic Modification of Barrow-Tsallis Entropy Can Alleviate the Hubble Tension,

    M. Yarahmadi and A. Salehi, “Generalized Logarithmic Modification of Barrow-Tsallis Entropy Can Alleviate the Hubble Tension,”Phys. Dark Univ., 10.1016/j.dark.2025.101923 (2025)

    arXiv 2025

  46. [48]

    Beyond the Local Universe: Impacts of Scalar field Coupled to Non-Relativistic Neutrinos on Bulk Flow,

    M. Yarahmadi, “Beyond the Local Universe: Impacts of Scalar field Coupled to Non-Relativistic Neutrinos on Bulk Flow,” Astropart. Phys., 10.1016/j.astropartphys.2025.103110 (2025)

    arXiv 2025

  47. [49]

    A., 2009, J

    Stern, D., Jimenez, R., Verde, L., Kamionkowski, M., and Stanford, S. A., 2009, J. Cosmol. Astropart. Phys., 2, 008, arXiv:0907.3149

    Pith/arXiv arXiv 2009

  48. [50]

    L., Loubser, S

    Ratsimbazafy, A. L., Loubser, S. I., Crawford, S. M., Cress, C. M., Bassett, B. A., Nichol, R. C., and V¨ ais¨ anen, P., 2017, Mon. Not. R. Astron. Soc., 467, 3239, arXiv:1702.00418

    Pith/arXiv arXiv 2017

  49. [51]

    Zhang, C., Zhang, H., Yuan, S., Liu, S., Zhang, T.-J., and Sun, Y.-C., 2014, Res. Astron. Astrophys., 14, 1221-1233, arXiv:1207.4541

    Pith/arXiv arXiv 2014

  50. [52]

    Borghi, N., Moresco, M., and Cimatti, A., 2022, Astrophys. J. Lett., 928, L4, arXiv:2110.04304 [astro-ph.CO]

    Pith/arXiv arXiv 2022

  51. [53]

    2024, MNRAS,Pages 3055–3067, https://doi.org/10.1093/mnras/stae2257

    M Yarahmadi, A Salehi. 2024, MNRAS,Pages 3055–3067, https://doi.org/10.1093/mnras/stae2257

    work page doi:10.1093/mnras/stae2257 2024

  52. [54]

    Distance priors from Planck final release,

    L. Chen, Q.-G. Huang, and K. Wang, “Distance priors from Planck final release,”J. Cosmol. Astropart. Phys.02(2019) 028, doi:10.1088/1475-7516/2019/02/028

    doi:10.1088/1475-7516/2019/02/028 2019

  53. [56]

    DESI 2024 Results III: Cosmological Implications of BAO Measurements,

    DESI Collaboration, “DESI 2024 Results III: Cosmological Implications of BAO Measurements,” arXiv:2503.14739 [astro- ph.CO] (2025)

    Pith/arXiv arXiv 2024

  54. [57]

    et al., arXiv:1907.12875 [astro-ph.CO]

    Aghanim, N. et al., arXiv:1907.12875 [astro-ph.CO]

    Pith/arXiv arXiv 1907

  55. [58]

    Astrophys

    Aghanim, N.; Akrami, Y.; Ashdown, M.; Aumont, J.; Baccigalupi, C.; Ballardini, M.; Banday, A.J.; Barreiro, R.; Bartolo, N.; Basak, S., 2020, Astron. Astrophys. 2020, 641, A8

    2020

  56. [59]

    Eur.Phys.J.C 84 (2024) 4, 443

    M Yarahmadi, A Salehi. Eur.Phys.J.C 84 (2024) 4, 443

    2024

  57. [60]

    DESI DR2 results. II. Measurements of baryon acoustic oscillations and cosmological constraints,

    DESI Collaboration, “DESI DR2 results. II. Measurements of baryon acoustic oscillations and cosmological constraints,” 28 Phys. Rev. D112, 083515 (2025), arXiv:2503.14738 [astro-ph.CO]

    Pith/arXiv arXiv 2025

  58. [61]

    Dynamical Dark Energy in the Crosshairs: A Joint Analysis with DESI, Pantheon plus, and TDCOSMO Constraints,

    T. Liu, X. Li, and J. Wang, “Dynamical Dark Energy in the Crosshairs: A Joint Analysis with DESI, Pantheon plus, and TDCOSMO Constraints,” arXiv:2504.21373 [astro-ph.CO] (2025)

    arXiv 2025

  59. [62]

    How Complex is Dark Energy? A Bayesian Analysis of CPL Extensions with Recent DESI BAO Measurements,

    M. Malekjani, S. Pourojaghi, and Z. Davari, “How Complex is Dark Energy? A Bayesian Analysis of CPL Extensions with Recent DESI BAO Measurements,” arXiv:2511.18657 [astro-ph.CO] (2025)

    Pith/arXiv arXiv 2025

  60. [63]

    Using non-DESI data to confirm and strengthen the DESI 2024 spatially flat w0waCDM cosmological parametrization result,

    C.-G. Park, J. de Cruz P´ erez, and B. Ratra, “Using non-DESI data to confirm and strengthen the DESI 2024 spatially flat w0waCDM cosmological parametrization result,”Phys. Rev. D110, 123533 (2024)

    2024

  61. [64]

    M Yarahmadi.2024, Physics of dark universe.Volume 47,101733,https://doi.org/10.1016/j.dark.2024.101733

    arXiv 2024

  62. [65]

    Yarahmadi et al

    M. Yarahmadi et al. ,2025,physics of dark universe, https://doi.org/10.1016/j.dark.2025.101824

    arXiv 2025

  63. [66]

    Cosmology with massive neutrinos coupled to dark energy,

    A. W. Brookfield, C. van de Bruck, D. F. Mota, and D. Tocchini-Valentini, “Cosmology with massive neutrinos coupled to dark energy,”Phys. Rev. Lett.96, 061301 (2006), arXiv:astro-ph/0503349

    Pith/arXiv arXiv 2006

  64. [67]

    Neutrinos in Cosmology,

    J. Lesgourgues and L. Verde, “Neutrinos in Cosmology,” in Particle Data Group Review (2023)

    2023

  65. [68]

    Status of neutrino properties and future prospects: Cosmological and astrophysical con- straints,

    M. Lattanzi and M. Gerbino, “Status of neutrino properties and future prospects: Cosmological and astrophysical con- straints,” arXiv:1712.07109 [astro-ph.CO] (2017)

    Pith/arXiv arXiv 2017

  66. [69]

    Effects of neutrino mass and asymmetry on cosmological structure formation,

    Z. Zeng, S. Yeung, and M.-C. Chu, “Effects of neutrino mass and asymmetry on cosmological structure formation,” arXiv:1808.00357 [astro-ph.CO] (2018)

    Pith/arXiv arXiv 2018

  67. [70]

    Can Chameleon fields be the source of both the dark energy dipole and the CMB dipole?

    M. Yarahmadi and A. Salehi, “Can Chameleon fields be the source of both the dark energy dipole and the CMB dipole?” Astrophys. J.964, 87 (2025)

    2025

  68. [71]

    Freedman, B.F

    W.L. Freedman, B.F. Madore, T. Hoyt et al.,arXiv:2002.01550(2020)

    Pith/arXiv arXiv 2002

  69. [72]

    Huang, A.G

    C.D. Huang, A.G. Riess, W. Yuan et al.,Astrophys. J.,889, 5, arXiv:1908.10883 (2019)

    Pith/arXiv arXiv 1908

  70. [73]

    Wong, S.H

    K.C. Wong, S.H. Suyu, G.C.-F. Chen et al.,Mon. Not. Roy. Astron. Soc., doi:10.1093/mnras/stz3094, arXiv:1907.04869 (2019)

    Pith/arXiv arXiv 1907

  71. [74]

    Baxter and B.D

    E.J. Baxter and B.D. Sherwin,arXiv:2007.04007(2020)

    Pith/arXiv arXiv 2007

  72. [75]

    Riess, S

    A.G. Riess, S. Casertano, W. Yuan, L.M. Macri, and D. Scolnic,Astrophys J.,876, 85 (2019) [arXiv:1903.07603]

    Pith/arXiv arXiv 2019

  73. [76]

    Wei and F

    J.-J. Wei and F. Melia,Astrophys.J.,897, 127, arXiv:2005.10422 (2020)

    Pith/arXiv arXiv 2005

  74. [77]

    Verde, T

    L. Verde, T. Treu, and A.G. Riess,Nature Astronomy,3, 891 (2019), arXiv:1907.10625

    Pith/arXiv arXiv 2019

  75. [78]

    Towards a Machine Learning Solution for Hubble Tension: Physics-Informed Neural Network (PINN) Analysis of Tsallis Holographic Dark Energy in Presence of Neutrinos,

    M. Yarahmadi and A. Salehi, “Towards a Machine Learning Solution for Hubble Tension: Physics-Informed Neural Network (PINN) Analysis of Tsallis Holographic Dark Energy in Presence of Neutrinos,”Eur. Phys. J. C1301, 85 (2025)

    2025

  76. [79]

    Can dark energy evolve to the phantom?

    A. Vikman, “Can dark energy evolve to the phantom?”Phys. Rev. D71, 023515 (2005) [arXiv:astro-ph/0407107]

    Pith/arXiv arXiv 2005

  77. [80]

    The future evolution and finite-time singularities inf(R) gravity,

    S. Nojiri and S. D. Odintsov, “The future evolution and finite-time singularities inf(R) gravity,”Phys. Rev. D74, 086005 (2006) [arXiv:hep-th/0601213]

    Pith/arXiv arXiv 2006

  78. [81]

    Unified cosmic history in modified gravity: fromF(R) theory to Lorentz non-invariant models,

    S. Nojiri and S. D. Odintsov, “Unified cosmic history in modified gravity: fromF(R) theory to Lorentz non-invariant models,”Phys. Rept.505, 59–144 (2011) [arXiv:1011.0544]

    Pith/arXiv arXiv 2011

  79. [82]

    Models off(R) cosmic acceleration that evade solar-system tests,

    W. Hu and I. Sawicki, “Models off(R) cosmic acceleration that evade solar-system tests,”Phys. Rev. D76, 064004 (2007) [arXiv:0705.1158]. 29

    Pith/arXiv arXiv 2007