Cosmology and Nongalactic Astrophysics · arXiv:2611.16407 v2 · submitted 12 Nov 2026
Minimal vacuum histories beyond ΛCDM: model comparison within VCDM
The paper says
“Recent cosmological observations motivate alternatives to ΛCDM with evolving dark energy. We compare three such models, each introducing one additional cosmological parameter: F42, with equation of state w(a) = w0/√a, and two smooth sign-changing effective dark-energy profiles, Ag and Qui.”
more of the abstract
All models are implemented within VCDM, which determines the background and linear perturbations without introducing an additional propagating gravitational degree of freedom. We use CMB observations from SPT, Planck, and ACT, combined with DESI DR2 distance measurements, with and without the DESY5 Dovekie supernova sample. All three alternatives improve the best-fit likelihood relative to ΛCDM. When supernovae are included, the Akaike information criterion favors F42, while Ag and Qui remain nearly indistinguishable from ΛCDM on this criterion. The Bayesian evidence, however, does not establish a robust preference among the three alternatives. Our results show that improved fits to current observations can arise from physically different dark-energy histories, motivating complementary measurements of expansion and structure growth to distinguish them.
from the abstract
Pith says
Machine review · deepseek-v4-flash · 13 Nov 2026 · of version 1 of the paper; version 2 is newer · what was read was not recorded · not checked by a person
It can be wrong. Challenge this review or reply to it.
Holds up. Solid, transparent model comparison that honestly reports it can't distinguish the alternatives — the nonlinear prescription is a known caveat, not a hidden flaw. This line is wrong
2 major objections · 5 minor
How often objections from reviews like this one are right: not yet measured
Not checked: figures, code, what the references say.
The paper says, at §II.A, §VI
The review did not quote this passage.
A classifier read this sentence of the paper as acknowledging the concern (probability 0.90):
“We use the same HMcode-2020 prescription [66] for all models, but its accuracy for these VCDM histories has not been independently established.”
Pith says: Objection 1 of 2 · major
The nonlinear prescription HMcode-2020, which affects CMB lensing and matter power spectrum predictions entering all likelihoods, is acknowledged as not independently validated for VCDM backgrounds (§II.A: 'Using the same prescription does not, however, establish its validity for nonlinear VCDM structure formation.').
more of objection 1
The evidence differences among models are ΔlnZ ≈ 1–2 with uncertainties ≈ 0.6. A systematic error of order 0.6 from inaccurate nonlinear modeling could therefore alter the relative ranking. While applying the same prescription to all models may partially cancel biases, the direction and magnitude of residual systematics for VCDM histories are unknown. The cross-check in Appendix D changes multiple analysis choices simultaneously (BAO, low-ℓ polarization, nonlinear prescription), so it does not isolate this effect. The authors should discuss the expected size of such systematics or, if feasible, perform a robustness test that varies the nonlinear prescription (e.g., comparing HMcode-2020 with Halofit or a different nonlinear model) for all models to bound the potential impact on the evidence differences.
The paper says, at §IV.B, Table V, Appendix A
The review did not quote this passage.
A classifier read this sentence of the paper as acknowledging the concern (probability 0.90):
“This adjustment accounts for the excluded prior volume, but does not recover likelihood contributions from poten- tially valid points rejected because of numerical failures.”
Pith says: Objection 2 of 2 · major
The 'box' evidence estimates for Ag and Qui are significantly reduced relative to the native estimates because only ≈ 6–7% of the prior mass (f_M ≈ 0.07) is retained by the likelihood evaluation domain. The dominant reason is restrictions from the Commander and MUSE likelihood implementations.
more of objection 2
The paper correctly adjusts for this but assigns zero likelihood to the excluded region. The small diagnostic check (40 points per run, §Appendix A) identifies causes but does not bound the missing likelihood contribution. If some excluded points give non-negligible likelihoods, the true evidence would differ from the box estimate. Given that the evidence differences of interest are comparable to the quoted uncertainties, the systematic uncertainty from this prior-volume exclusion should be more prominently discussed. A sensitivity analysis (e.g., assuming a range of possible likelihood values for the excluded region) would help gauge robustness.
5 minor objections
The paper says, at §I, table captions
Pith says: minor objection 1 of 5
The labels 'Λs Ag' and 'Λs Qui' appear in figures and Table II, but the models are not standard 'Λs' profiles. To avoid confusion, the text should explain this naming or simply use 'Ag' and 'Qui' consistently.
The paper says, at §IV.C, Eq. (17)
Pith says: minor objection 2 of 5
The marginalized statistic Q_SN in Eq. (17) includes a logarithmic term (ln(1^T C^{-1} 1 / (2π))). Please clarify explicitly that this is the fully marginalized likelihood (the log of the Bayesian evidence for the SN data after marginalizing over the magnitude offset), as opposed to a profiling statistic. The statement that it is used consistently for all models is good, but the nature of the quantity should be unambiguous.
show 3 more
The paper says, at §III.B
Pith says: minor objection 3 of 5
The compressed DESI DR2 Lyα BAO–Alcock–Paczyński constraints used here may assume a fiducial cosmology for the compression. The paper notes this assumption (§VI) but a brief reference to how this was checked (e.g., the range of validity in model space) would be useful.
The paper says, at Table II
Pith says: minor objection 4 of 5
The column 'z× or z†' is interpreted differently for F42 (phantom crossing) and Ag/Qui (transition center). For Ag, the text states that z† does not mark the density sign change because of the Gaussian bump. A brief note in the table caption or footnote would prevent misinterpretation.
The paper says, at Appendix A, Table VI
Pith says: minor objection 5 of 5
The weight diagnostic N_w = 1/∑ w_i^2 is reported but not interpreted. A sentence indicating that values around 1500–1700 out of ~11000 samples suggest reasonable effective sample sizes for the PolyChord run would be helpful (or a comment on the expected range).
What would settle it
Re-run the full PolyChord analysis for one alternative model (e.g., F42) with D0 data, replacing HMcode-2020 with a different nonlinear prescription (e.g., Halofit with the Mead et al. 2015 recalibration or the Takahashi et al. 2012 prescription), while keeping all other settings (DESI Lyα BAO-AP compression, low-ell likelihood, etc.) identical. Compare the resulting best-fit chi-squared and Bayesian evidence against the HMcode-2020 values. If Δχ² changes by more than 1 or ΔlnZ changes by more than the PolyChord uncertainty (~0.6), the model comparison is sensitive to the nonlinear prescription choice.
from the skeptic's reading
How the review went
4 readings by deepseek-v4-flash · 13 Nov 2026
- Reader holds up
- Skeptic agrees
- Editor wrote the verdict above
- Referee minor revision: 2 major objections, 5 minor
Not run on this paper: the simulated authors' defence.
Pith did not check this review's objections against the paper.
The review in full
The review in brief, machine-written
The paper does what it says: it compares three one-parameter dark-energy histories (F42, Ag, Qui) within the VCDM framework using the latest CMB-SPA, DESI DR2 (with Lyα AP), and DESY5 Dovekie data. All three fit better than ΛCDM by χ², but the Bayesian evidence doesn't pick a winner. The AIC favors F42 when supernovae are included, but the evidence differences are within the noise. The authors are refreshingly honest about the limitations — the HMcode-2020 nonlinear prescription isn't validated for these backgrounds, the evidence depends on prior conventions for the excluded likelihood region, and they never claim a detection.
The main novelty is the simultaneous comparison in a common theoretical framework (VCDM) with the newest data. The individual models were introduced earlier, but this kind of head-to-head benchmark is useful. The analysis is methodologically sound: nested sampling, matched priors, separate annealing for best-fits, and clear reporting of both AIC and Bayes factors. The supplementary cross-check with Cobaya and different settings (Halofit, low-ℓ EE) gives some reassurance that the posteriors are robust.
Soft spots: The HMcode-2020 concern is real but not buried. The authors state it explicitly in Section VI and in the paper's own sentence [b54]. The stress-test worry — that systematic shifts of order ΔlnZ ~ 0.6 could change rankings — is valid, but since the evidence differences are already comparable to the numerical uncertainty, the conclusion (no robust preference) still stands. A larger issue is that the Ag and Qui posteriors pile up at the lower prior boundary a†=0.25, which is the early-transition limit approaching ΛCDM. The paper acknowledges this, but the prior choice drives the evidence comparison. Extending the prior would change the Bayes factors, as they discuss in Appendix C. Finally, the compressed DESI Lyα AP distances assume the distance extraction is valid for these cosmologies — another caveat they flag.
Overall, this is a solid, transparent analysis that will be useful for the cosmology community working on DESI and Euclid forecasts. It doesn't overclaim. I'd send it to peer review without hesitation.
The strongest objection
The paper relies on HMcode-2020 for nonlinear corrections to the matter power spectrum and CMB lensing predictions. Although the same prescription is applied to all models, its calibration to ΛCDM-like simulations is not validated for the VCDM backgrounds (F42, Ag, Qui) which have different expansion histories and linear growth factors. The paper explicitly states this limitation (Section VI, sentence [b239]) and that it 'does not establish its validity for nonlinear VCDM structure formation' (Section II). The likelihoods for CMB lensing (SPT-3G MUSE, ACT–Planck) and DESI BAO (which include the Alcock–Paczyński signal) depend on the predicted nonlinear matter power spectrum. If HMcode-2020 systematically misestimates the nonlinear enhancement for these models, the chi-squared values and hence the AIC and evidence differences could be biased. The cross-check in Appendix D for Qui alone replaces both Halofit and multiple other settings, so it does not isolate the nonlinear-prescription effect. The evidence differences among models are of order ΔlnZ ≈ 1–2, and a systematic error of ∼0.6 (comparable to the quoted uncertainty) could alter the ranking. This makes the nonlinear prescription a load-bearing assumption.
The reader identified the HMcode-2020 accuracy as the weakest assumption. I agree: it is an acknowledged but untested assumption that directly affects the predicted observables entering the likelihoods. The paper's careful cross-check for Qui is not sufficient because it changes multiple analysis choices simultaneously and does not cover F42 or Ag. However, the authors transparently state this limitation and do not overclaim robustness. Their main conclusion—that multiple histories improve the fit but are not robustly distinguished—remains supported because the evidence differences are small with overlapping uncertainties. A dedicated sensitivity test would strengthen the paper but is not required to retain the ACCEPT verdict. Therefore, I recommend UNCHANGED.
The referee's report
If the result holds, it demonstrates that current cosmological data cannot robustly distinguish between qualitatively different dark-energy histories—a continuously evolving positive density (F42) and sign-changing profiles (Ag/Qui)—at matched parameter count. This is a valuable finding for the community because it clarifies the limits of present constraints and highlights the need for complementary measurements (e.g., Euclid) that probe both expansion and growth. The analysis is thorough: it uses state-of-the-art data, implements all models within a single theoretical framework (VCDM) to avoid ad hoc perturbation assumptions, and is transparent about limitations. The paper's honest reporting of systematic uncertainties, prior volume effects, and numerical checks sets a good standard for model comparison studies.
Pith's reading of the paper (a machine's summary, not the authors')
This paper compares three one-parameter extensions of ΛCDM—F42, Ag, and Qui—implemented within the VCDM theory, which shares general relativity's two gravitational degrees of freedom. Using CMB data from SPT, Planck, and ACT, plus DESI DR2 distances, the authors find that all three alternatives improve the best-fit likelihood over ΛCDM. When supernovae are included, the Akaike information criterion favors F42, while Ag and Qui remain nearly indistinguishable from ΛCDM on that criterion. However, the Bayesian evidence does not establish a robust preference among the three alternatives. The results show that improved fits to current data can arise from physically different dark-energy histories, motivating future measurements of expansion and structure growth.
What carries the argument. The key machinery is the VCDM gravity theory, which allows a chosen homogeneous expansion history to be realized through an auxiliary potential while propagating only the two tensor modes of general relativity. This avoids introducing a phenomenological dark-energy sound speed. The background histories are F42 (with equation of state w(a)=w0/√a), Ag (hyperbolic-tangent transition with a Gaussian bump), and Qui (monotonic hyperbolic-tangent transition), each adding a single parameter to ΛCDM.
Core claim. The central finding is that minimal one-parameter dark-energy histories—a continuously evolving equation of state (F42) or a sign-changing effective density (Ag and Qui)—all yield better fits to combined CMB, BAO, and supernova data than ΛCDM, yet current data cannot robustly distinguish among them. F42 favors a lower Hubble constant and higher matter density, while Ag and Qui favor the opposite. All three models predict higher S8 than ΛCDM. The VCDM framework ensures consistent perturbation equations without extra degrees of freedom, so the differences are purely in the background expansion history. The Bayesian evidence, after prior-volume adjustments, shows central values that mildly favor the alternatives over ΛCDM, with log Bayes factors between 0.6 and 2.3, but the uncertainties prevent a decisive ranking.
The premise it rests on. The accuracy of the HMcode-2020 nonlinear prescription for structure formation in VCDM histories has not been independently established; this directly affects the predicted matter power spectrum and CMB lensing signals that enter the likelihoods.
If the paper is right (Pith's inferences, not the authors')
F42 predicts a later phantom-crossing redshift (z×≈0.42) when supernovae are included, compared to earlier crossing without them.
Ag and Qui favor a transition near z†≈2.7 and push H0 above 68.6 km/s/Mpc and Ωm below 0.304, shifting the inferred Hubble constant and matter density relative to ΛCDM.
All three models increase S8 by about 0.01 relative to ΛCDM, affecting structure-growth inferences.
Future Euclid lensing and clustering data can distinguish these histories because their expansion and growth histories differ beyond the distance measures used here.
The similar parameter constraints for Ag and Qui imply that the localized density enhancement in Ag is not significantly detected by current data.
If the HMcode-2020 nonlinear prescription is inaccurate for VCDM, the reported likelihood improvements could be biased; independent N‑body simulations for these histories would test that directly.
The posterior peak of Ag and Qui at a†≈0.27 lies near the prior boundary; extending the uniform prior below 0.25 could move the effective density toward the ΛCDM limit and change the evidence ranking.
The VCDM perturbation equations predict different growth histories even for identical H(z) if the auxiliary potential differs; redshift‑space distortion measurements from DESI or Euclid could break the degeneracy between Ag and Qui despite their similar background fits.
Check this review
Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.
How often reviews like this one are right
Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.
How often objections from reviews like this one are right: not yet measured
Reviews by people
This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.
The paper
Abstract
Recent cosmological observations motivate alternatives to ΛCDM with evolving dark energy. We compare three such models, each introducing one additional cosmological parameter: F42, with equation of state w(a) = w0/√a, and two smooth sign-changing effective dark-energy profiles, Ag and Qui. All models are implemented within VCDM, which determines the background and linear perturbations without introducing an additional propagating gravitational degree of freedom. We use CMB observations from SPT, Planck, and ACT, combined with DESI DR2 distance measurements, with and without the DESY5 Dovekie supernova sample. All three alternatives improve the best-fit likelihood relative to ΛCDM. When supernovae are included, the Akaike information criterion favors F42, while Ag and Qui remain nearly indistinguishable from ΛCDM on this criterion. The Bayesian evidence, however, does not establish a robust preference among the three alternatives. Our results show that improved fits to current observations can arise from physically different dark-energy histories, motivating complementary measurements of expansion and structure growth to distinguish them.
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.
abstract compiled · 2026-11-13
The authors' main claim, as the classifier picked it out 1.00
“Our results show that improved fits to current observations can arise from physically different dark-energy histories, motivating complementary measurements of expansion and structure growth to distinguish them.”
How the main claim is stated: asserted 0.97
What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.90 · human participants 0.01 · controlled comparison 0.87 · randomized assignment 0.01 · states quantitative result 0.34 · compares with prior results 0.88 · mentions code or data availability 0.02
The abstract, sentence by sentence
This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.
| Sentence | Role | The authors' own work | States the main conclusion | Strength |
|---|---|---|---|---|
| Recent cosmological observations motivate alternatives to $\Lambda$CDM with evolving dark energy. | background 1.00 | 0.04 | 0.02 | asserted 0.88 |
| We compare three such models, each introducing one additional cosmological parameter: F42, with equation of state $w(a)=w_0/\sqrt{a}$, and two smooth sign-changing effective dark-energy profiles, Ag and Qui. | contribution 0.84 | 0.82 | 0.03 | asserted 0.98 |
| All models are implemented within VCDM, which determines the background and linear perturbations without introducing an additional propagating gravitational degree of freedom. | method 0.99 | 0.29 | 0.03 | asserted 1.00 |
| We use CMB observations from SPT, Planck, and ACT, combined with DESI DR2 distance measurements, with and without the DESY5 Dovekie supernova sample. | method 1.00 | 0.75 | 0.02 | asserted 0.92 |
| All three alternatives improve the best-fit likelihood relative to $\Lambda$CDM. | result 1.00 | 0.91 | 0.15 | asserted 1.00 |
| When supernovae are included, the Akaike information criterion favors F42, while Ag and Qui remain nearly indistinguishable from $\Lambda$CDM on this criterion. | result 1.00 | 0.94 | 0.15 | asserted 1.00 |
| The Bayesian evidence, however, does not establish a robust preference among the three alternatives. | result 0.98 | 0.92 | 0.14 | asserted 0.97 |
| Our results show that improved fits to current observations can arise from physically different dark-energy histories, motivating complementary measurements of expansion and structure growth to distinguish them. | interpretation 0.97 | 0.91 | 0.89 | asserted 0.97 |
What kind of paper the classifier reports this is: empirical study 0.49 · evidence from analysis of existing data 1.00
What was examined, and what was not
Sources read
- arXiv listing abstract v1
Not examined
- full text
- figures and tables
- references and citing sentences
- relations to other papers
Levels run
- P abstract
Not run
- S sections
- T body sentences
- C citations
- R cross-paper relations
How the labels were made
Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-13: 1 request, 6,244 input tokens.
Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON
Figures (4)
Figures are the authors', reproduced from the paper.
Topics: Dark Energy, Dark Energy
Challenges and discussion
Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.
Come back by email
A short list of the week's newly reviewed papers. Nothing is sent until you confirm from your inbox, and every email has a link that stops it. No account.
No comments yet.
Sign in to comment