Pith. sign in

REVIEW 2 major objections 4 minor 49 references

Spatial anisotropies from long wavelength scalar and tensor modes

T0 review · 2 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The leading physical effect of superhorizon scalar and tensor perturbations in a flat adiabatic universe is a metric locally equivalent to one of three Bianchi cosmologies: types I, V, and IX, depending on the sign of the scalar-induced…

desk verdict The gradient-expansion derivation of the effective metric (35) and the Bianchi VII no-go are solid, but the Bianchi V correspondence only works for a measure-zero class of tensor modes, so the paper's main claim overreaches. read the letter →

arxiv 2412.15181 v2 pith:5YPQOYN5 submitted 2024-12-19 gr-qc

Jorge Noreña , Thiago S. Pereira , Sean K. Reynolds This is my paper · ORCID
classification gr-qc MSC 83F05 PACS 98.80.-k04.20.-q
keywords BianchicosmologiessuperhorizonperturbationsgradientexpansionspatialcurvaturetensoradiabaticFLRWuniversecosmologicalperturbationtheory
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

Superhorizon fluctuations in a flat, matter-dominated FLRW universe are usually treated as locally unobservable, but this paper shows their leading physical effect has a precise geometric shape. Working to first order in perturbations and second order in a gradient expansion, with adiabatic matter in comoving gauge, the authors derive an effective metric consisting of a curved FLRW background with a homogeneous tensor perturbation superposed. This metric is locally indistinguishable from Bianchi type I, V, or IX, depending on whether the scalar-induced spatial curvature $k=-\frac{2}{3}\nabla^2 R$ is zero, negative, or positive. The result gives a concrete meaning to the intuition that superhorizon modes 'look like' curved or anisotropic universes, and it rules out Bianchi VII$_0$ and VII$_h$ in this setting. A sympathetic reader should care because it ties the perturbative treatment of the early universe to the Bianchi classification of homogeneous cosmologies.

What carries the argument

The load-bearing object is the effective metric of Eq. (35), $ds^2=-dt^2+b(t)^2[(1-\frac{k}{2}x^2)\delta_{ij}+\gamma_{ij}(t)]dx^idx^j$, reached by eliminating the pure-gauge leading and subleading orders of the gradient expansion and passing to synchronous coordinates. The derivation uses the comoving-gauge ADM metric, the Hamiltonian and momentum constraints (which fix the lapse perturbation as $\delta N=\dot{R}/H$ and make the shift order $q/H$), and the identification of the Laplacian of the curvature perturbation with a spatial curvature $k$. The key structural step is that the traceless part of the scalar Hessian drops out of the Einstein equations in the absence of anisotropic stress, leaving only the trace, which becomes $k$, and the homogeneous tensor mode $\gamma_{ij}(t)$, which plays the role of shear. The comparison with Bianchi models is then carried out at the level of the linearized Einstein tensor and of the Killing-vector algebras of the Bianchi spaces.

What would settle it

Compute the space-space Einstein equations for the metric (24) including a nonzero anisotropic stress $\Pi_{ij}$ in the matter sector: the traceless part of $A_{kl}$ in Eq. (34) would no longer vanish, so the metric could not be reduced to (35) and the Bianchi I/V/IX identification would break, even for infinitesimal $\Pi_{ij}$. Observationally, a dataset at horizon scales that requires Bianchi VII$_0$ or VII$_h$ symmetry while the matter is adiabatic and stress-free would also falsify the claim.

Watch

Extended reading notes

Core claim

The paper's central claim is that the combination of long-wavelength scalar and tensor adiabatic modes in a flat FLRW universe is locally equivalent, at leading order in gradients, to three specific Bianchi cosmologies: type I when the induced curvature vanishes, type V when it is negative, and type IX when it is positive. Scalar perturbations act through the trace of their Hessian, which renormalises the spatial curvature, while the off-diagonal Hessian terms are ineffective because adiabatic perfect-fluid matter sources no anisotropic stress. Long-wavelength tensor perturbations supply the homogeneous shear of the model. The authors prove the equivalence both by matching the linearized Einstein equations for the effective metric with those of the Bianchi models and by constructing explicit invariant bases of Killing vectors, following and extending the known result that Bianchi IX is a gravitational wave of maximal wavelength on a positively curved background. They also show that Bianchi VII$_0$ and VII$_h$ cannot arise as adiabatic perturbations of a flat FLRW universe with a non-tilted perfect fluid.

Load-bearing premise

The result assumes the matter perturbations are adiabatic and in comoving gauge, so the off-diagonal Hessian terms of the scalar perturbation produce no shear and only its trace survives as spatial curvature; if matter had anisotropic stress or non-adiabatic pressure, this reduction would fail.

Editorial extensions

If this is right

  • A flat universe with adiabatic superhorizon perturbations is locally a Bianchi I, V, or IX universe, so horizon-scale anisotropy is not independent of the curvature induced by scalar modes.
  • Bianchi VII$_0$ and VII$_h$ are excluded for adiabatic perfect-fluid perturbations, meaning any observational evidence for those symmetries would require non-adiabatic pressure, anisotropic stress, or a tilted fluid.
  • The scalar curvature-mimicry result, $k=-\frac{2}{3}\nabla^2 R$, is recovered as the trace sector of the general effective metric, independent of the tensor modes.
  • The effective geometry provides a framework for computing an observational anisotropy floor at horizon scales, a direction the paper explicitly proposes as follow-up work.

Reading between the lines

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

  • If a tiny anisotropic stress is present on superhorizon scales, the traceless part of the scalar Hessian would source shear and the metric would not reduce to Eq. (35); computing $A_{kl}$ with $\Pi_{ij}\neq0$ would show exactly how the Bianchi I/V/IX equivalence degrades.
  • Applied to a matter content that is not a simple adiabatic perfect fluid—for example, with entropy perturbations or a vector field—the same gradient expansion may produce Bianchi VII$_h$ or other classes, because the no-go relies specifically on the adiabatic condition.
  • The sign correlation between local curvature and shear type (positive $k$ only with Bianchi IX, negative only with V, zero with I) is testable in principle with horizon-scale observations, since measuring both would distinguish the three geometries.
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

2 major / 4 minor

Summary. The paper considers superhorizon scalar and tensor perturbations in a matter-dominated flat FLRW universe, working to first order in perturbations and second order in a gradient expansion. It shows that the constant and linear-gradient parts of the curvature perturbation R and the tensor perturbation γ_ij are pure gauge, and that the leading physical effect of the scalar Laplacian is a local spatial curvature k = -2/3 ∇²R. Including long-wavelength tensor modes, the authors derive an effective metric (35) and argue that it is locally equivalent to Bianchi types I, V, and IX depending on the sign of k, with the tensor perturbation providing the shear. They further argue that Bianchi VII_0 and VII_h cannot arise in this adiabatic, non-tilted perfect-fluid setting. The derivation uses a series of coordinate transformations to simplify the metric, a comparison of the linearized Einstein tensors, and an explicit construction of Killing vectors and invariant bases.

Significance. If correct, the paper provides a clean and useful dictionary between long-wavelength cosmological perturbations and a subset of homogeneous anisotropic cosmologies, with potential implications for interpreting large-angle anomalies and for the separate-universe approach. The paper is self-contained and its main steps are amenable to independent verification; the gauge arguments in Section II and the no-go results for VII_0 and VII_h are clear. The derivation is honest about its assumptions (adiabatic perturbations, non-tilted perfect fluid, matter domination). The central limitation is that the Bianchi V case appears to hold only for a restricted class of tensor configurations, which is not stated in the abstract.

major comments (2)
  1. [§IV.B, Eq. (40) and the k<0 bullet] The Bianchi V matching requires a nonzero vector a_i satisfying a_j dotγ_ij = 0 when n_ij=0. For a generic traceless symmetric 3×3 matrix dotγ_ij, no nonzero null eigenvector exists, so the only solution is a_i=0, which corresponds to Bianchi I rather than V. The text's statement that one can take a_i in the direction to which γ_ij is transverse is therefore valid only for the degenerate class of tensor configurations for which dotγ_ij has a null eigenvector (e.g., a single plane-wave mode). For a generic superposition of long-wavelength tensor modes, condition (40) fails and the claimed equivalence with Bianchi V is not established. Since the abstract and conclusions state that generic adiabatic scalar and tensor perturbations in the k<0 case yield Bianchi V, this is a load-bearing restriction that must be stated explicitly and justified.
  2. [§IV.C] The invariant-basis construction for Bianchi V shows that any constant γ_rs gives a metric with Bianchi V symmetry, but this does not demonstrate that the metric satisfies the Einstein equations with a non-tilted perfect fluid. The dynamical check is the Einstein-tensor comparison in Section IV.B, and since condition (40) generically fails, the metric (35) with k<0 is not a non-tilted perfect-fluid Bianchi V solution for generic γ_ij. The logical role of the metric comparison in Section IV.C relative to the Einstein-tensor comparison should be clarified, and the precise class of tensor configurations for which the V equivalence holds should be stated.
minor comments (4)
  1. [Eq. (27)] The standard Riemann normal coordinate expansion is g^(3)_ij = δ_ij - (1/3) R_ikjl x^k x^l + O(x^3); please verify the sign and factor in Eq. (27) and check whether it propagates consistently through Eqs. (29)–(31).
  2. [Eq. (34)] The trace decomposition in Eq. (34) appears to double-count the trace term; please clarify the indexing of δG^i_j and the steps that lead to A_kl = -(k/2)δ_kl from the vanishing of anisotropic stress.
  3. [Eq. (40) and §IV.B] In Eq. (40), '3aaj' appears to be a typo for '3a_j', and the subsequent bullet refers to a_i being orthogonal to γ_ij, while the condition actually involves dotγ_ij; the required property is that dotγ_ij has a null eigenvector.
  4. [Title page] The word 'tensor' is split as 'tenso r' in the title line; please fix this typo.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the effective metric is derived from perturbation theory and matched to Bianchi geometries via independent Einstein-equation and Killing comparisons; self-citations are non-load-bearing.

full rationale

The derivation is self-contained. Starting from the standard comoving-gauge perturbed metric (1), the paper removes leading-order constant perturbations and constant gradients by explicit coordinate transformations (Secs. II.A-II.B), rederives the known equivalence between the scalar Laplacian and spatial curvature k = -2/3 nabla^2 R (Sec. II.C and Appendix B), and obtains the effective metric (35) by imposing the absence of anisotropic stress through Eq. (34). The Bianchi types are then identified by comparing the resulting Einstein tensor with the linearized Bianchi equations (20)-(22) and by explicitly constructing invariant Killing bases (Sec. IV.C). None of these steps uses the Bianchi conclusion as an input: the metric (35) is not assumed to be Bianchi, and the Bianchi matching imposes additional conditions (Eqs. 39-41) that are not automatic. The citations to King (1991) and Pereira & Pitrou (2019) are prior independent calculations; the latter, despite author overlap, is used only for the standard linearized Bianchi Einstein equations, which are also cited to Pontzen & Challinor [25], and does not presuppose the present equivalence. There are no fitted parameters renamed as predictions and no uniqueness theorem imported from the authors. The reviewer concern that condition (40) for Bianchi V is not generically satisfiable by arbitrary superpositions of long-wavelength tensor modes is a substantive correctness/assumption question, not a circularity: the derivation does not force Eq. (40) by construction.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The derivation rests on the standard assumptions of cosmological perturbation theory: adiabatic superhorizon perturbations in comoving gauge, a matter-dominated perfect-fluid background, and a gradient expansion truncated at second order. No free parameters are fitted, and no new dynamical entities are introduced. The effective curvature k is derived from the perturbation R, not assumed.

assumptions (6)
  • domain assumption Perturbations are adiabatic and treated in comoving gauge, sourced by a single-field inflationary model.
    Used throughout Section II and IV to fix lapse and shift and to drop off-diagonal Hessian terms of the scalar perturbation; stated in the Introduction and in Section IV.A.
  • domain assumption The background universe is matter-dominated (w=0).
    Appendix B uses the matter-dominated density relation (Eq. B2) to check that the effective scale factor b(t) obeys the curved Friedmann equations; also ensures tensor modes are constant on superhorizon scales.
  • domain assumption Superhorizon modes justify a gradient expansion truncated at second order in x.
    Expansion in Eq. (3); all higher-order terms are neglected. The equivalence is therefore local and at leading nontrivial order.
  • domain assumption The background spatial curvature vanishes; the scalar perturbation's Laplacian induces an effective curvature k = -(2/3) nabla^2 R.
    Section II.C establishes k = -(2/3) nabla^2 R; this k enters the metric (35) and the Bianchi comparison (Eq. 39).
  • domain assumption The Einstein equations are linearized over perturbations, and the comparison with Bianchi models uses the linearized equations for gamma_ij.
    Comparison of the Einstein tensors in Section IV.B relies on linearization in gamma_ij and in the curvature k; the compatibility conditions (39)-(41) follow from this linearization.
  • domain assumption The fluid is a perfect, non-tilted fluid with no anisotropic stress.
    The perfect non-tilted fluid with no anisotropic stress is necessary to set the traceless part of the space-space Einstein tensor to zero, yielding A_kl = -(k/2) delta_kl, and to match the Bianchi source terms.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Spatial anisotropies from long wavelength scalar and tensor modes." pith.science (2026). https://pith.science/paper/5YPQOYN5

@misc{pith2026241215181,
  author       = {Pith},
  title        = {Pith review of: Spatial anisotropies from long wavelength scalar and tensor modes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5YPQOYN5}},
  note         = {Machine review of arXiv:2412.15181}
}
read the original abstract

We investigate the dominant physical effects of superhorizon fluctuations in a flat FLRW universe, focusing on whether the combined evolution of scalar and tensor adiabatic modes in the near-horizon regime could lead to geometries beyond those predicted by the conventional separate-universe approach. Assuming a matter-dominated universe and working to first order in perturbations but second order in a gradient expansion, we identify modes that are either pure gauge or unsourced, making them observationally irrelevant. This allows us to derive an effective metric that preserves the spatial symmetries of three well-known Bianchi cosmologies, namely, types I, V, and IX. In this framework, scalar perturbations induce spatial curvature, while the shear arises from long-wavelength tensor perturbations.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 25 canonical work pages

  1. [1]

    we eliminated the gradient through the coordinate trans- formation of the previous section

    takes the form ds2 = − ( 1 + 2 ˙R H ) dt2 +a2(t) ( 1 + 2R(t) + 1 3 x2∇2R ) dx2, (12) where we have ignored the time-dependence of ∂iR and ∇2R (since they would be of higher order in q/H ), and 1 Note that this is also the transformation to the Riemann coor - dinates on the constant t hypersurface at this order. we eliminated the gradient through the coord...

    work page

  2. [2]

    The starting point is the metric in the ADM formalism 6 ds2 = −N 2dt2 +a2(t)hij(dxi +N idt)(dxj +N jdt), (A1) wherehij = (1+2 R)δij +γij

    from the constraint equations. The starting point is the metric in the ADM formalism 6 ds2 = −N 2dt2 +a2(t)hij(dxi +N idt)(dxj +N jdt), (A1) wherehij = (1+2 R)δij +γij. The momentum and Hamil- tonian constraints are (3)R +K 2 −KijK ij = 16πGρ, (A2) Di(K ij −a−2hijK) = −8πGNT 0j, (A3) where the extrinsic curvature is Kij = 1 2aN (∂t(a2hij) − DiNj −DjNi) an...

    work page

  3. [3]

    Second, the gradient expansion also intro- duces anisotropic derivatives of R which do not appear in the metric above

    (and thus the lapse ( 2)). Second, the gradient expansion also intro- duces anisotropic derivatives of R which do not appear in the metric above. We shall come back to this issue later, and for the moment we just focus on the diagonal part. Therefore, the perturbed metric (

    work page

  4. [4]

    is indistinguishable from a subset of Bianchi metrics with small curvature and anisotropies, namely, models I, V and IX (see Table I). A. Metric After performing the coordinate transformations that eliminate the leading and sub-leading orders in the gra- dient expansion (that is, a constant R andγij, plus their gradients), and now including both off-diagon...

    work page

  5. [5]

    Planck, P. A. R. Ade et al. , Astron. Astrophys. 594, A16 (2016), [1506.07135]

    arXiv 2016

  6. [6]

    For the closed k >0 case, the solution to this question can be found in [ 24]

    is indistinguishable from such a homogeneous but anisotropic universe. For the closed k >0 case, the solution to this question can be found in [ 24]. We reproduce their result here expanding to second order in x. The Killing vectors of Bianchi IX are ξ± r = 1 2 (Tr ± √ kRr), (48) where one chooses either the plus or minus sign for the three translational ...

    work page

  7. [7]

    P. K. Aluri et al. , Class. Quant. Grav. 40, 094001 (2023), [2207.05765]

    arXiv 2023

  8. [8]

    Bull et al

    P. Bull et al. , Phys. Dark Univ. 12, 56 (2016), [1512.05356]

    arXiv 2016

Show all 49 references
  1. [9]

    Marozzi and J.-P

    G. Marozzi and J.-P. Uzan, Phys. Rev. D 86, 063528 (2012), [1206.4887]

    2012 arXiv

  2. [10]

    Anselmi et al

    S. Anselmi et al. , JCAP 02, 049 (2023), [2207.06547]

    2023 arXiv

  3. [11]

    This is easily fixed with a time transformation t ↦→t +ξ0 with ˙ξ0 = − ˙R H

    and ( 12) is that the latter is not in a synchronous coordinate system. This is easily fixed with a time transformation t ↦→t +ξ0 with ˙ξ0 = − ˙R H . The metric then becomes ds2 = −dt2 +b2(t) ( 1 − 1 2kx2 ) dx2, (13) where b(t) ≡a(t +ξ0)(1 + R(t)), k ≡ − 2 3 ∇2R. (14) This has ...

  4. [12]

    Aghanim et al

    Planck, N. Aghanim et al. , Astron. Astrophys. 641, A6 (2020), [1807.06209], [Erratum: Astron.Astrophys. 652, C4 (2021)]

    2020 arXiv

  5. [13]

    are those of a curved FLR W. Note that the time-dependence of b(t) = a(t +ξ0)(1 + R(t)) is fixed by the time-dependence of a,R,ξ0 so we want to check that this time dependence solves the curved FLR W Friedmann equations. Taking a time derivative of the effective scale factor b(t...

  6. [14]

    Alam et al

    eBOSS, S. Alam et al. , Phys. Rev. D 103, 083533 (2021), [2007.08991]

    2021 arXiv

  7. [15]

    D. J. Schwarz, C. J. Copi, D. Huterer and G. D. Starkman, Class. Quant. Grav. 33, 184001 (2016), [1510.07929]

    2016 arXiv

  8. [16]

    Abdalla et al

    E. Abdalla et al. , JHEAp 34, 49 (2022), [2203.06142]

    2022 arXiv

  9. [17]

    T. P. Waterhouse and J. P. Zibin, 0804.1771

    arXiv

  10. [18]

    C. D. Leonard, P. Bull and R. Allison, Phys. Rev. D 94, 023502 (2016), [1604.01410]

    2016 arXiv

  11. [19]

    T. S. Pereira and C. Pitrou, Phys. Rev. D 100, 123534 (2019), [1909.13688]

    2019 arXiv

  12. [20]

    R. H. Brandenberger and C. S. Lam, hep-th/0407048

    arXiv

  13. [21]

    Geshnizjani, D

    G. Geshnizjani, D. J. H. Chung and N. Afshordi, Phys. Rev. D 72, 023517 (2005), [astro-ph/0503553]

    2005 arXiv

  14. [22]

    C. M. Hirata and U. Seljak, Phys. Rev. D 72, 083501 (2005), [astro-ph/0503582]

    2005 arXiv

  15. [23]

    Baldauf, U

    T. Baldauf, U. Seljak, L. Senatore and M. Zaldarriaga, JCAP 10, 031 (2011), [1106.5507]

    2011 arXiv

  16. [24]

    Kleban and M

    M. Kleban and M. Schillo, JCAP 06, 029 (2012), [1202.5037]

    2012 arXiv

  17. [25]

    Knox, Phys

    L. Knox, Phys. Rev. D 73, 023503 (2006), [astro- ph/0503405]

    2006

  18. [26]

    J. M. Maldacena, JHEP 05, 013 (2003), [astro- ph/0210603]

    2003

  19. [27]

    J. M. Bardeen, Phys. Rev. D 22, 1882 (1980)

    1980

  20. [28]

    Tian et al

    C. Tian et al. , Phys. Rev. D 103, 083513 (2021), [2010.07274]

    2021 arXiv

  21. [29]

    Blachier, P

    B. Blachier, P. Auclair, C. Ringeval and V. Vennin, Phys . Rev. D 108, 123510 (2023), [2302.14530]

    2023 arXiv

  22. [30]

    Talebian-Ashkezari, N

    A. Talebian-Ashkezari, N. Ahmadi and A. A. Abolhasani, JCAP 03, 001 (2018), [1609.05893]

    2018 arXiv

  23. [31]

    Adamek, R

    J. Adamek, R. Durrer and V. Tansella, JCAP 01, 024 (2016), [1510.01566]

    2016 arXiv

  24. [32]

    O. H. Marcori, C. Pitrou, J.-P. Uzan and T. S. Pereira, Phys. Rev. D 98, 023517 (2018), [1805.12121]

    2018 arXiv

  25. [33]

    D. H. King, Phys. Rev. D 44, 2356 (1991)

    1991

  26. [34]

    Pontzen and A

    A. Pontzen and A. Challinor, Class. Quant. Grav. 28, 185007 (2011), [1009.3935]

    2011 arXiv

  27. [35]

    Since anisotropy is small, we argue that this shows that (

    to arbitrary order in k (but still expanding in gradients) such that it is a Bianchi metric. Since anisotropy is small, we argue that this shows that (

  28. [36]

    G. F. Ellis, R. Maartens and M. A. MacCallum, Rela- tivistic cosmology (Cambridge University Press, 2012)

    2012

  29. [37]

    D. H. Lyth, Phys. Rev. D 31, 1792 (1985)

    1985

  30. [38]

    Bordin, P

    L. Bordin, P. Creminelli, M. Mirbabayi and J. Nore˜ na, JCAP 09, 041 (2016), [1605.08424]

    2016 arXiv

  31. [39]

    L. Dai, E. Pajer and F. Schmidt, JCAP 11, 043 (2015), [1502.02011]

    2015 arXiv

  32. [40]

    Creminelli, J

    P. Creminelli, J. Nore˜ na and M. Simonovi´ c, JCAP 07, 052 (2012), [1203.4595]

    2012 arXiv

  33. [41]

    Hinterbichler, L

    K. Hinterbichler, L. Hui and J. Khoury, JCAP 08, 017 (2012), [1203.6351]

    2012 arXiv

  34. [42]

    Creminelli and M

    P. Creminelli and M. Zaldarriaga, JCAP 10, 006 (2004), [astro-ph/0407059]

    2004 arXiv

  35. [43]

    Similarly, if ξ− r are chosen as the Killing vectors, ξ+ r form an invariant basis and satisfy a similar equation

    and ( 44) that δrsξ− r ξ− s = ( 1 + k 2 x2 ) δij, (50) 7 so these vectors indeed form an invariant basis at this order: er =ξ− r . Similarly, if ξ− r are chosen as the Killing vectors, ξ+ r form an invariant basis and satisfy a similar equation. Now we can explicitly build a m...

    2024

  36. [44]

    Weinberg, Cosmology (OUP Oxford, 2008)

    S. Weinberg, Cosmology (OUP Oxford, 2008)

    2008

  37. [45]

    Grøn and S

    Ø. Grøn and S. Hervik, Einstein ’s general theory of rel- ativity: with modern applications in cosmology (Springer Science & Business Media, 2007)

    2007

  38. [46]

    Plebanski and A

    J. Plebanski and A. Krasinski, An introduction to general relativity and cosmology (Cambridge University Press, 2024)

    2024

  39. [47]

    Weinberg, Gravitation and Cosmology: Principles and Applications of the General Theory of Relativity (John Wiley & Sons, New York, 1972)

    S. Weinberg, Gravitation and Cosmology: Principles and Applications of the General Theory of Relativity (John Wiley & Sons, New York, 1972)

    1972

  40. [48]

    Hinterbichler, L

    K. Hinterbichler, L. Hui and J. Khoury, JCAP 01, 039 (2014), [1304.5527]

    2014 arXiv

  41. [49]

    Mirbabayi and M

    M. Mirbabayi and M. Zaldarriaga, JCAP 03, 025 (2015), [1409.6317]. 6 Note that we have factorized the scale factor. This means tha t quantities defined in terms of derivatives of hij receive no back- ground contribution

    2015 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.