Pith. sign in

REVIEW 3 major objections 5 minor 36 references

Tensor-Induced Backreaction in Ultra-Slow-Roll Inflation

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Tensor-induced backreaction during an ultra-slow-roll phase lowers the curvature power spectrum peak from ~10^-3 to below 4×10^-6, making primordial black hole production negligible in the benchmark models.

desk verdict Tensor backreaction in USR is a real and interesting effect, but the headline suppression number rests on a partial resummation that is not controlled, so treat the quantitative claim as conditional. read the letter →

arxiv 2608.07264 v1 pith:BEDOVFET submitted 2026-08-07 astro-ph.CO gr-qchep-th

Alessandro Ciaiolo , Giovanni Marozzi This is my paper · ORCID
classification astro-ph.COgr-qchep-th MSC 83F0583C25 PACS 98.80.Cq
keywords ultra-slow-rollinflationtensor-inducedbackreactionprimordialblackholesscalarpowerspectrumgauge-invariantaveragingself-consistentresummationMukhanov-Sasakiequationfine-tuning
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 argues that tensor modes, usually ignored in backreaction studies, act back on the inflationary background at second order and become the dominant correction exactly in the ultra-slow-roll phase where primordial black hole production is expected. The effect limits how small the first slow-roll parameter ε can become, and since the curvature power spectrum is amplified when ε is tiny, capping ε caps the spectrum. In two benchmark realizations of a single-field model, the authors find the peak falls from about 2×$10^{-3}$ and 7×$10^{-3}$ to roughly 4×$10^{-6}$, making PBH production negligible. A self-consistent resummation beyond fixed-background perturbation theory confirms that the effect remains important even when the corrected background is used. The same mechanism strongly reduces the fine-tuning of the peak amplitude with respect to the potential's bump location.

What carries the argument

The machinery is the gauge-invariant averaging of backreaction with the inflaton as the clock, yielding effective Hubble and slow-roll parameters in Eqs. (11)–(12). The tensor contribution enters through $B=-\langle \dot{\bar\psi}_t^{(2)}\rangle - \frac{1}{24}\langle h_{ij}\dot h^{ij}\rangle$, with correlators evaluated in the long-wavelength limit as $\langle h_{ij}\dot h^{ij}\rangle = H^3/(M_{\rm Pl}^2\pi^2)$ and $\langle \dot{\bar\psi}_t^{(2)}\rangle = \frac{13}{64\pi^2} H^3/(M_{\rm Pl}^2\varepsilon)$. The coefficient $13/64$ rests on the identity in Eq. (36). To go beyond fixed-background perturbation theory, the authors define a self-consistent Hubble parameter $H_{\rm sc}$ via Eq. (16), derive effective slow-roll parameters $\varepsilon_{\rm sc}$, $\eta_{\rm sc}$ by differentiating $H_{\rm sc}$, and solve the closed system of Eqs. (18)–(20). The corrected power spectrum is obtained by solving the Mukhanov–Sasaki equation with these self-consistent background quantities.

What would settle it

The central claim is settled by computing the full second-order backreaction system without relying on the identity (36) and with the inflaton equation of motion consistently corrected; if the curvature peak stays near $10^{-3}$ rather than dropping below $4\times10^{-6}$, the suppression is not robust.

Watch

Extended reading notes

Core claim

The central claim, as the authors would state it, is that in single-field inflation with a transient ultra-slow-roll phase, second-order tensor-induced backreaction is not a negligible correction but a leading effect. It curbs the fall of ε, shortens the non-attractor phase, and suppresses the comoving curvature power spectrum peak by about three orders of magnitude, to slightly below 4×$10^{-6}$ in the benchmark models. Consequently, primordial black hole production in these models becomes negligible, and the fine-tuning associated with the peak amplitude is substantially reduced. The authors emphasize that this is a partial resummation: genuine higher-order tensor contributions and corrections to the inflaton equation of motion and to the perturbation evolution equations are not included, so they stop short of declaring a no-go theorem.

Load-bearing premise

The quantitative suppression rests on an unproven identity, Eq. (36), which fixes the coefficient 13/64, and on the assumption that the backreaction is fully captured by replacing $H$ and $\varepsilon$ with the self-consistent $H_{\rm sc}$ and $\varepsilon_{\rm sc}$ only in the source term while leaving the inflaton equation and the perturbation equations unchanged.

Editorial extensions

If this is right

  • The peak of the scalar power spectrum in the benchmark ultra-slow-roll models drops from about 2–7×10^-3 to slightly below 4×10^-6 once tensor backreaction is included.
  • Primordial black hole production, which depends exponentially on the peak amplitude, becomes negligible in these models.
  • The non-attractor phase lasts about 1.1 e-folds instead of 2.2–2.4, and the two benchmark realizations converge to nearly the same peak value, so the model is much less sensitive to the bump-location parameter.
  • The self-consistent treatment shows the backreaction does not disappear when evaluated on the corrected background, ruling out a fixed-background artifact at this order.

Reading between the lines

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

  • Editorial extension: if the suppression survives the removal of the not-included corrections (inflaton equation of motion, higher-order tensor modes, pure scalar backreaction), the paper's conclusion effectively becomes a no-go for PBH production from single-field ultra-slow-roll phases, which is stronger than what the authors formally claim.
  • Editorial extension: the $1/\varepsilon^2$ growth of the $\varepsilon$ and $\eta$ corrections is generic in non-attractor phases, so the same suppression should appear in other USR realizations; testing a second benchmark potential would show whether $4\times10^{-6}$ is a universal cap.
  • Editorial extension: because the backreaction also affects modes that leave the horizon before the ultra-slow-roll phase, the shape of the power spectrum on nearby scales changes; this could leave a distinctive signature in the induced gravitational wave background even when PBHs are absent.
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.

Referee Report

3 major / 5 minor

Summary. The paper studies second-order tensor-induced backreaction during a transient ultra-slow-roll phase of single-field inflation. Using a covariant, observer-dependent averaging formalism, the authors derive corrections to the effective Hubble parameter and slow-roll parameters, and then introduce a 'self-consistent' scheme in which the backreaction source is evaluated on a corrected background. Applying this to two realizations of a benchmark potential, they find that the curvature power spectrum peak is reduced from about 2e-3 and 7e-3 to slightly below 4e-6, making primordial black hole production negligible, and they argue that the fine-tuning of the peak amplitude is strongly reduced. The paper explicitly lists several limitations of the procedure, including the neglect of scalar backreaction and the assumption that the perturbation equations keep their form.

Significance. If the central quantitative claim is correct, the result is significant for PBH phenomenology: it would imply that tensor-induced backreaction can qualitatively change the predicted PBH abundance in a broad class of USR models and simultaneously relax the notorious fine-tuning of the peak amplitude. The paper also has clear strengths: the benchmark parameters are fixed by CMB normalization and by the requirement of realizing a USR phase, not by fitting to the backreaction result; the two model realizations differ only by a small shift in one potential parameter; and the authors are unusually explicit about the limitations of their resummation. However, the headline numbers rest on an unproven identity and on a partial resummation whose control parameter does not directly bound the slow-roll quantities that set the peak height, so the significance will be fully realized only after those points are secured.

major comments (3)
  1. [Beyond the perturbative approach for ultra-slow-roll, Eqs. (16)-(20)] The self-consistent scheme is only a partial resummation and its control parameter is not established for the quantities that determine the peak. In Eq. (16), H_sc^2 is defined using the bare H^2 and the bare ε on the right-hand side, and Eqs. (18)-(19) still evolve H_sc and ε_sc with respect to the bare e-fold time N while the background Hubble flow and the inflaton equation of motion are kept unchanged. The stated condition |δH^2|/H^2 of order 10^-3 bounds only the correction to H^2, not the O(1) modifications of ε_sc and η_sc during the non-attractor phase. Because the claimed suppression of Δ_R^2 from about 10^-3 to below 4×10^-6 is obtained with this scheme, the quantitative conclusion is not yet robust; please either derive a controlled expansion parameter for the resummation or quantify the sensitivity of the peak amplitude to alternative, equally plausible partial resummations (for example, replacing H by H_sc also in the definition of e-fold time and in the Friedmann and Klein-Gordon equations).
  2. [End Matter, evaluation of ⟨˙ψ_t^(2)⟩ and ⟨ḣ_ij h_ij⟩, Eq. (36)] The identity ⟨(∂i∂k/∇^2)(h_km ḣ_mi)⟩ = -⟨h_ij ḣ_ij⟩ is stated without proof, and it fixes the numerical coefficient 13/64 in Eq. (39) that propagates into all backreaction formulas. The identity is not obvious: in Fourier space, with transverse traceless polarization tensors, contractions of the type k_i e_mi and k_k e_km vanish, so the non-zero value and the sign require a careful treatment of the mode expansion and of the ∇^{-2} operator. Please provide a derivation or an explicit reference, and ideally an independent check of the final coefficient in Eq. (14).
  3. [End Matter, 'Scalar backreaction for a comoving observer'] The neglect of scalar backreaction is not supported by the cited theorem in the regime studied here. The text states that the no-scalar-backreaction result of Refs. [23,25] applies when the comoving curvature perturbation is conserved on super-Hubble scales, and then notes that in USR it is not conserved. The subsequent argument is a physical expectation rather than a calculation, and the claim that the backreaction-modified dynamics would in any case reduce scalar backreaction assumes part of the effect under examination. Since scalar perturbations carry comparably large power in these models, a quantitative estimate or a clear statement that the main results are conditional on this assumption is needed.
minor comments (5)
  1. [Introduction, paragraph on SR and USR] The sentence 'standard slow-roll (SR) regime requires all slow-roll parameters to be small' is clear, but the later phrase 'limiting the suppression of ε' is ambiguous; consider rewording to 'limiting the decrease of ε' or 'preventing ε from becoming as small as in the unperturbed case.'
  2. [Figure 1 and Figure 2] The figures are informative, but the text should explicitly state the units of H^2 (M_Pl^2) in the axes and clarify whether the self-consistent quantities are plotted against N or N_sc; the current caption says they are plotted against N_sc, but Eq. (18) evolves in N.
  3. [After Eq. (15)] The statement that the perturbative corrections to εeff and ηeff become non-perturbative when ε = O(H/M_Pl) should be made more precise: the relevant criterion is that the coefficients multiplying 1/ε^2 in Eq. (15) become of order unity, and this should be stated explicitly to avoid the impression that the breakdown is tied to a specific numerical value of H/M_Pl.
  4. [End Matter, Eqs. (41)-(42)] The fine-tuning measure is defined as an absolute logarithmic derivative in Eq. (41), but Eq. (42) evaluates it as a finite difference between two realizations; please state explicitly that Eq. (42) approximates the derivative and comment on the sensitivity of the reported reduction to the finite step in φ_d.
  5. [General] There are a few typographical issues, for example 'becames each other coupled' should read 'become coupled to each other', and some sentences in the End Matter are run-on; a careful proofread would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the peak suppression follows from a parameter-free backreaction computation, not from a fit to the target spectrum.

full rationale

The central claim is not circular because the backreaction coefficients entering Eq. (15) are computed from the correlators of Eq. (14), which are evaluated in the End Matter (Eqs. (30)-(40)) and do not use the scalar power-spectrum peak as an input. The benchmark parameters are fixed by CMB normalization (A*_R = 2.1e-9, phi*(N*) = 3 M_Pl) and by the choice of Gaussian-bump parameters; the two values of phi_d are chosen before backreaction is included, and no parameter is adjusted to make Delta^2_Rsc reach 4e-6. Eq. (16) is an ansatz for a partial resummation, and the paper explicitly lists its limitations (no extension to the inflaton equation of motion, unchanged perturbation equations, neglected higher-order tensor and scalar backreaction); this makes the numerical peak model-dependent, but not equivalent to its input. The reliance on prior work by the authors ([18,19,20,23,35]) supplies the gauge-invariant averaging framework and the second-order constraint equations, but the new non-attractor result and the 13/64 coefficient are derived in this paper (Eqs. (31)-(39)); the cited results are not the target prediction. The unproven identity Eq. (36) is a mathematical step affecting the coefficient, i.e. a correctness risk, not a circular reduction.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

No new particles, forces, or entities are introduced. The free parameters are benchmark model inputs, not fitting parameters for the backreaction effect. The main assumptions concern the averaging scheme, the long-wavelength limit, an unproven correlator identity, and the ad hoc self-consistent resummation.

free parameters (1)
  • Benchmark potential parameters (V0, A, sigma, phi_d1, phi_d2) = V0 fixed by A_R*=2.1e-9 at pivot; A=1.17e-3; sigma=1.59e-2 M_Pl; phi_d1=2.18801 M_Pl; phi_d2=2.18801*1.00006 M_Pl
    These are chosen to realize a USR phase and match CMB observations, not fitted to the backreaction result. They are model inputs for the illustrative benchmark.
assumptions (5)
  • domain assumption The gauge-invariant averaging scheme of [18,19] is the correct observable definition of backreaction.
    Used to define H_eff and the effective slow-roll parameters in Eqs. (3)-(5).
  • domain assumption Second-order perturbation theory in the long-wavelength limit captures the leading tensor-induced backreaction, with vector modes kinematically suppressed.
    Stated in the main text; vector modes are neglected because they are kinematically suppressed during inflation.
  • ad hoc to paper The identity in Eq. (36), <∂i∂k/∇^2(h_km dot h_mi)> = -<h_ij dot h_ij>, is correct.
    No derivation is shown. This identity sets the 13/64 coefficient in Eq. (14) and is load-bearing for the quantitative result.
  • ad hoc to paper The self-consistent replacement H -> H_sc and epsilon -> epsilon_sc in the backreaction source (Eq. 16) is a valid partial resummation.
    The scheme is introduced by hand and checked only by the smallness of δH^2. The inflaton equation of motion and the functional form of the perturbation equations are left unchanged, as acknowledged in the Conclusions.
  • domain assumption Purely scalar backreaction is negligible for a comoving observer even though the comoving curvature perturbation is not conserved in ultra-slow-roll.
    The formal proof in [23,25] does not apply in USR; the paper substitutes a physical argument that the inflaton is the only source. This is a reasoned assumption, not a proven result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tensor-Induced Backreaction in Ultra-Slow-Roll Inflation." pith.science (2026). https://pith.science/paper/BEDOVFET

@misc{pith2026260807264,
  author       = {Pith},
  title        = {Pith review of: Tensor-Induced Backreaction in Ultra-Slow-Roll Inflation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BEDOVFET}},
  note         = {Machine review of arXiv:2608.07264}
}
abstract

We discuss the impact of tensor-induced backreaction at second order in perturbation theory on primordial black holes production in single-field inflationary models, where the comoving curvature power spectrum is amplified by a transient ultra-slow-roll phase. Although the leading perturbative correction is mostly negligible in standard slow-roll evolution, it becomes important in non-attractor regimes, going in the direction of limiting the suppression of the first slow-roll parameter $\varepsilon$ and reducing the enhancement of the scalar power spectrum. We then go beyond the standard perturbative analysis by considering the background parameters as effective quantities with evolution self consistently determined by a closed system of backreaction equations. This analysis, which leads to a partial resummation of the effect beyond the perturbative treatment, confirms the limitation of the suppression of $\varepsilon$ by consistently showing how the effect remains important even after accounting for the effective evolution of the background in the evaluation of the backreaction itself. Finally, we discuss how, taking into account the tensor-induced backreaction, the fine-tuning associated with the value of the peak amplitude of the scalar power spectrum is strongly reduced.

Figures

Figures reproduced from arXiv: 2608.07264 by the authors.

Figure 1
Figure 1. Evolution of H2 and H2 sc in panel (a), of ε and εsc in panel (b), and of η and ηsc in panel (c). Solid black curves are the first background model realization, and dot-dashed blue curves are the corresponding self-consistent quantities. Dashed black curves are the second background model real￾ization, and dotted blue curves are the corresponding self￾consistent quantities. for both cases. Therefore, the tensor back… view at source ↗
Figure 2
Figure 2. We show ∆2 R and ∆2 R sc in function of the e-folds for which the mode k exit the horizon. The solid black curve is ∆2 R for the first background model realization, and the dot￾dashed blue curve is the corresponding ∆2 R sc. The dashed black curve is ∆2 R for the second background model realiza￾tion, and the dotted blue curve is the corresponding ∆2 R sc. ture perturbation, d 2Rk dN2 + (3 − ε + η) dRk dN +  k aH 2… view at source ↗
Figure 3
Figure 3. The black curve is the fine-tuning measure [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 8 canonical work pages

  1. [1]

    A. H. Guth, Phys. Rev. D23, 347 (1981)

    1981

  2. [2]

    A. D. Linde, Phys. Lett. B108, 389 (1982)

    1982

  3. [3]

    V. F. Mukhanov and G. V. Chibisov, JETP Lett.33, 532 (1981)

    1981

  4. [4]

    A. A. Starobinsky, JETP Lett.30, 682 (1979)

    1979

  5. [5]

    Akramiet al.(Planck), Astron

    Y. Akramiet al.(Planck), Astron. Astrophys.641, A10 (2020), arXiv:1807.06211 [astro-ph.CO]

    arXiv 2020

  6. [6]

    W. H. Kinney, Phys. Rev. D72, 023515 (2005), arXiv:gr- qc/0503017. 6

    arXiv 2005

  7. [7]

    Martin, H

    J. Martin, H. Motohashi, and T. Suyama, Phys. Rev. D 87, 023514 (2013), arXiv:1211.0083 [astro-ph.CO]

    arXiv 2013

  8. [8]

    Dimopoulos, Phys

    K. Dimopoulos, Phys. Lett. B775, 262 (2017), arXiv:1707.05644 [hep-ph]

    arXiv 2017

Show all 36 references
  1. [9]

    Garcia-Bellido and E

    J. Garcia-Bellido and E. Ruiz Morales, Phys. Dark Univ. 18, 47 (2017), arXiv:1702.03901 [astro-ph.CO]

    2017 arXiv

  2. [10]

    Di and Y

    H. Di and Y. Gong, JCAP07, 007, arXiv:1707.09578 [astro-ph.CO]

    arXiv

  3. [11]

    J. E. Gammalet al.(LISA Cosmology Working Group), JCAP05, 062, arXiv:2501.11320 [astro-ph.CO]

    arXiv

  4. [12]

    S. S. Mishra and V. Sahni, JCAP04, 007, arXiv:1911.00057 [gr-qc]

    1911 arXiv

  5. [13]

    A. D. Gow, C.T. Byrnes, P.S. Cole, andS. Young, JCAP 02, 002, arXiv:2008.03289 [astro-ph.CO]

    2008 arXiv

  6. [14]

    P. S. Cole, A. D. Gow, C. T. Byrnes, and S. P. Patil, JCAP08, 031, arXiv:2304.01997 [astro-ph.CO]

    arXiv

  7. [15]

    Frolovsky and S

    D. Frolovsky and S. V. Ketov, Universe9, 294 (2023), arXiv:2304.12558 [astro-ph.CO]

    2023 arXiv

  8. [16]

    S. R. Geller, W. Qin, E. McDonough, and D. I. Kaiser, Phys. Rev. D106, 063535 (2022), arXiv:2205.04471 [hep- th]

    2022 arXiv

  9. [17]

    A. J. Iovino and A. Riotto, JCAP07, 047, arXiv:2512.19668 [astro-ph.CO]

    arXiv

  10. [18]

    Gasperini, G

    M. Gasperini, G. Marozzi, and G. Veneziano, JCAP03, 011, arXiv:0901.1303 [gr-qc]

    arXiv

  11. [19]

    Gasperini, G

    M. Gasperini, G. Marozzi, and G. Veneziano, JCAP02, 009, arXiv:0912.3244 [gr-qc]

    arXiv

  12. [20]

    Marozzi and G

    G. Marozzi and G. P. Vacca, Phys. Rev. D90, 043532 (2014), arXiv:1405.3933 [gr-qc]

    2014 arXiv

  13. [21]

    Buchert, Gen

    T. Buchert, Gen. Rel. Grav.32, 105 (2000), arXiv:gr- qc/9906015

    2000

  14. [22]

    Buchert, Gen

    T. Buchert, Gen. Rel. Grav.33, 1381 (2001), arXiv:gr- qc/0102049

    2001

  15. [23]

    Finelli, G

    F. Finelli, G. Marozzi, G. P. Vacca, and G. Venturi, Phys. Rev. Lett.106, 121304 (2011), arXiv:1101.1051 [gr-qc]

    2011 arXiv

  16. [24]

    Marozzi, JCAP01, 012, arXiv:1011.4921 [gr-qc]

    G. Marozzi, JCAP01, 012, arXiv:1011.4921 [gr-qc]

    arXiv

  17. [25]

    Marozzi and G

    G. Marozzi and G. P. Vacca, Phys. Rev. D88, 027302 (2013), arXiv:1304.2291 [gr-qc]

    2013 arXiv

  18. [26]

    Marozzi and G

    G. Marozzi and G. P. Vacca, Class. Quant. Grav.29, 115007 (2012), arXiv:1108.1363 [gr-qc]

    2012 arXiv

  19. [27]

    Marozzi, G

    G. Marozzi, G. P. Vacca, and R. H. Brandenberger, JCAP02, 027, arXiv:1212.6029 [hep-th]

    arXiv

  20. [28]

    D. C. Galanti, P. Conzinu, G. Marozzi, and S. San- tos da Costa, Phys. Rev. D110, 123510 (2024), arXiv:2406.19960 [gr-qc]

    2024 arXiv

  21. [29]

    P. A. R. Adeet al.(BICEP, Keck), Phys. Rev. Lett.127, 151301 (2021), arXiv:2110.00483 [astro-ph.CO]

    2021

  22. [30]

    Cheng, D.-S

    S.-L. Cheng, D.-S. Lee, and K.-W. Ng, Phys. Lett. B827, 136956 (2022), arXiv:2106.09275 [astro-ph.CO]

    2022 arXiv

  23. [31]

    Caravano, G

    A. Caravano, G. Franciolini, and S. Renaux-Petel, Phys. Rev. D111, 063518 (2025), arXiv:2410.23942 [astro- ph.CO]

    2025 arXiv

  24. [32]

    Geshnizjani and N

    G. Geshnizjani and N. Afshordi, JCAP01, 011, arXiv:gr- qc/0405117

  25. [33]

    Pattison, V

    C. Pattison, V. Vennin, H. Assadullahi, and D. Wands, JCAP07, 031, arXiv:1905.06300 [astro-ph.CO]

    1905 arXiv

  26. [34]

    Geshnizjani and R

    G. Geshnizjani and R. Brandenberger, Phys. Rev. D66, 123507 (2002), arXiv:gr-qc/0204074

    2002 arXiv

  27. [35]

    Finelli, G

    F. Finelli, G. Marozzi, G. P. Vacca, and G. Venturi, Phys. Rev. D69, 123508 (2004), arXiv:gr-qc/0310086

    2004 arXiv

  28. [36]

    Finelli, G

    F. Finelli, G. Marozzi, A. A. Starobinsky, G. P. Vacca, and G. Venturi, Phys. Rev. D79, 044007 (2009), arXiv:0808.1786 [hep-th]. End Matter Scalar backreaction for a comoving observer.— Regarding the purely scalar contribution to cosmolog- ical backreaction in non-attractor re...

    2009 arXiv

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.