Pith. sign in

REVIEW 3 major objections 4 minor 3 cited by

This paper derives exact formulas for the primordial three-point function when curvature and isocurvature fields mix strongly, without expanding in the mixing strength.

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

2026-08-01 23:43 UTC pith:R5DHCFS4

load-bearing objection First analytic strong-mixing bispectrum, but the central u-integrals are formal as written — needs a regularization prescription before the formulas are well-defined. the 3 major comments →

arxiv 2607.15251 v1 pith:R5DHCFS4 submitted 2026-07-16 hep-th

New exact bispectrum shapes in multifield inflation

Lucas Pinol This is my paper
classification hep-th
keywords multifield inflationprimordial bispectrumquadratic mixingcosmological collider signalnon-perturbative resummationisocurvature fluctuationsequilateral shapeSchwinger-parameter integral
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.

This paper aims to compute the three-point function of primordial curvature fluctuations in multifield inflation without treating the quadratic mixing between curvature and isocurvature modes as a small parameter. It claims the first exact analytical bispectrum for the cubic time-derivative interaction, obtained by recasting the exact mixed linear modes as a single integral and proving that every scale-invariant tree-level bispectrum collapses to one Schwinger-parameter integral over pre-computable leg kernels. The resulting shapes interpolate from the standard single-field equilateral form at weak mixing to genuinely multifield shapes at strong mixing, where the squeezed limit contains an exponentially enhanced cosmological collider oscillation whose frequency is set by the dressed mass, μ_eff = sqrt(λ² + m²/H² − 9/4). A sympathetic reader would care because the strong-mixing regime, where the collider signal is largest, had previously been accessible only numerically.

Core claim

On the paper's own terms, the central result is Eq. (13): the exact bispectrum shape for the cubic time-derivative vertex at any dimensionless mixing strength λ, expressed as a single Schwinger-parameter integral over pre-computable leg kernels. In the squeezed limit this reduces to the closed form S ≃ N√κ A(μ_eff, λ) sin[μ_eff ln κ − δ(μ_eff, λ)], showing that the collider-clock frequency is exactly μ_eff = sqrt(λ² + m²/H² − 9/4) at every mixing strength, and that the amplitude is Boltzmann-enhanced by channel weights e^{aπλ/2}, growing like e^{2πλ} at fixed μ_eff. The weak-mixing λ≪1 limit reproduces the standard equilateral single-field result, while strong mixing produces shapes that dec

What carries the argument

The load-bearing object is the single-integral representation of the exact mixed mode functions: each curvature mode is a u-integral over a weight ω_a(u) of a single dressed plane wave, with ω_a(u) given by a hypergeometric function. The Schwinger parameter ξ separates the three external frequencies and factorizes the in-in time integral into leg kernels W_n^a(β), evaluated at rescaled arguments β_j = 2ξ e_j. Together with the channel weights e^{aπλ/2} and the boundary coefficients r_a, these pieces turn any scale-invariant tree-level one-vertex diagram into a one-dimensional ξ-integral over independent, pre-computable kernels.

Load-bearing premise

Everything rests on the single-integral representation (8)–(9) for the exact linear mixed mode functions, which is inherited from an earlier operator construction and verified numerically here rather than derived; if that representation is wrong, the leg kernels and all bispectrum formulas fail.

What would settle it

Compute the cubic bispectrum at (λ, μ_eff) = (2.5, 2.5) by direct numerical in-in integration of the coupled mode equations without using Eq. (8), and compare the shape over the full triangle with Eq. (13); agreement to numerical precision supports the claim, while any systematic discrepancy falsifies it. A lighter check: derive the O(λ²) squeezed amplitude of Eq. (18) directly from perturbative diagrams and compare.

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

If this is right

  • Every scale-invariant tree-level contact bispectrum, including those with isocurvature legs, reduces to the same single-integral representation; exchange diagrams become contact diagrams because the mixed fields no longer commute.
  • At weak mixing λ≲0.5 the shape correlates with the equilateral template at better than 0.99; at strong mixing it decorrelates, with the cosine crossing zero and reaching about −0.9, so equilateral templates miss the dominant phenomenology.
  • The squeezed-limit clock frequency is exactly μ_eff at any λ, matching earlier numerical and semi-analytical results, and the amplitude is exponentially enhanced relative to the perturbative λ² scaling.
  • The λ≪1 limit reproduces the known equilateral single-field bispectrum, and the new O(λ²) collider amplitude provides a check against future perturbative calculations.

Where Pith is reading between the lines

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

  • If these shapes are correct, existing CMB and large-scale-structure constraints built on equilateral or orthogonal templates would not capture strongly mixed models; a dedicated template search could place first bounds on λ≳1.
  • The method is cheap enough that full parameter scans over (λ, m, μ_eff) are feasible, which could identify the most observable regions before designing a future survey.
  • The exponential enhancement raises the question of loop corrections: if similar e^{2πλ} factors appear in higher-order diagrams, the loop expansion may require resummation, which the paper leaves open.
  • The paper's discussion of tachyonic bare masses at strong mixing suggests sub-Hubble instabilities that may leave distinctive imprints in mildly squeezed configurations; extending the analysis to that parameter space is a natural next step.

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 an analytical computation of a primordial bispectrum in a multifield inflation model where the quadratic mixing between curvature and isocurvature fluctuations is treated non-perturbatively. It recasts the exact linear solutions of Ref. [1] into single-integral representations, defines leg kernels, and shows that all scale-invariant tree-level single-vertex bispectra reduce to a Schwinger-parameter integral over these kernels. The method is applied to the pi_dot^3 interaction: at weak mixing the shape reduces to the equilateral one, while at strong mixing it acquires large, decorrelated multifield features and an exponentially enhanced cosmological collider signal with frequency set by an effective mass meff^2 = m^2 + rho^2. A closed-form squeezed-limit expression is obtained at arbitrary mixing strength, and several consistency checks are reported.

Significance. If the central formulas are made rigorous, this is a valuable technical advance: it would provide the first fully analytic handle on the strong-mixing regime of multifield inflation, where previous treatments were numerical or semi-analytical. The Schwinger-parameter reduction is elegant and general, the explicit squeezed-limit expression is a falsifiable prediction, and the effective-mass interpretation of the collider clock frequency is physically illuminating. The paper also supplies several internal consistency checks: the weak-mixing limit matches the standard equilateral shape, the O(lambda^2) collider scaling matches the expected perturbative structure, and the strong-mixing power-spectrum enhancement agrees with earlier numerical results. However, the validity of the entire construction depends on the integral representation of the linear mode functions, which is imported from Ref. [1] and is not independently derived in this manuscript.

major comments (3)
  1. [Eqs. (8)-(9), (12) and App. A] The integrals defining the mode functions and leg kernels are not convergent as ordinary improper integrals for real lambda != 0. Near u=0 the weight in Eq. (9) behaves as u^{-1 + i a lambda/2} / Gamma(i a lambda/2), so the integral over [0,epsilon] is proportional to epsilon^{i a lambda/2}/(i a lambda/2), which has no limit as epsilon -> 0. The statement below Eq. (12) that 'the integral is always convergent' is therefore not justified. Since the kernels (12) enter the central shape formula (13) and the squeezed-limit result (17), a regularization prescription (analytic continuation in lambda, contour deformation, or endpoint subtraction) must be supplied and shown to yield the correct Bunch-Davies normalization. Without this, Eqs. (13) and (17) are formal expressions.
  2. [App. A, Eq. (21)] The single-integral representation (8)-(9) is stated to follow from the operator representation of Ref. [1], but the derivation is not given; the text only says 'We have verified that our new integral representation reproduces exactly the linear solutions of Ref. [1]'. This is load-bearing because every subsequent kernel, shape, and squeezed-limit formula depends on this representation. In particular, the standard integral representation of the Tricomi function U(a,b,z) invoked in App. A is usually stated for Re a > 0, whereas here the relevant exponent has vanishing real part. The analytic continuation to Re a = 0 must be made explicit. The manuscript should either provide a self-contained derivation or state the precise domain and regularized definition.
  3. [App. B, last paragraph] The assertion that '[pi_c, sigma] != 0 even for interaction picture fields' is, as written, inconsistent with canonical equal-time commutation relations. Equal-time fields commute; what can be nonzero is the unequal-time commutator [pi_I(0), sigma_I(tau)] because the free Hamiltonian contains the mixing term -rho sigma p_pi. This distinction matters for evaluating in-in expectation values involving sigma operators. The conclusion that exchange diagrams become contact diagrams does not require a nonzero equal-time commutator; it follows from treating the mixing as part of the free Hamiltonian. Please clarify the statement and correct the reasoning.
minor comments (4)
  1. [Eq. (13)] The notation lambda_2 for the cubic coupling and lambda for the mixing strength is easy to confuse. In particular, Lambda_star = lambda_2^{-1/2} is introduced without a fully explicit relation to the action (1)-(3).
  2. [Eq. (18)] The O(lambda^2) squeezed limit is compared to a perturbative calculation that is said to be 'not available in the literature'. This weakens the check; it would be useful to show the two-insertion diagram explicitly or compare with a direct numerical in-in evaluation.
  3. [Regime of validity] Several statements about the sigma-pi^2 interactions and the companion paper [17] cannot be checked because [17] is not available. Please indicate which parts of the regime-of-validity discussion rely on unpublished work.
  4. [Figure 2] The discrete points in Figure 2 are drawn from Eq. (13), but the convergence/regularization procedure used for the numerical evaluation is not described. State the cutoff or continuation method used.

Circularity Check

0 steps flagged

No significant circularity; the central derivation is self-contained given externally cited linear solutions.

full rationale

The paper's central claim is the exact bispectrum shape (13), built from the mixed-system mode functions (8)-(9). Those mode functions are imported from Ref. [1], an independent group's preprint, not from the author's own prior work; the paper derives its own single-integral recasting in App. A and verifies it against Ref. [1]. This is external support, not a self-citation chain. The effective mass (5), meff^2 = m^2 + rho^2, is obtained by a canonical transformation of the quadratic Hamiltonian (4), not by fitting the collider frequency; the identification mu_eff = sqrt(lambda^2 + m^2/H^2 - 9/4) is an algebraic consequence of that definition and is checked against prior numerical/semi-analytical results, not used as an input. The reduction of all scale-invariant tree-level one-vertex diagrams to a single Schwinger integral is proved in App. B from standard in-in time integrals. The weak-mixing O(lambda^2) and strong-mixing limits are validated against independent perturbative and numerical calculations. No parameter is fitted to the quantity later called a prediction. The convergence/regularization status of the u-integrals in (8)-(9) for real lambda is a mathematical-rigor concern, not a circularity, since the derivation does not presuppose the bispectrum result.

Axiom & Free-Parameter Ledger

0 free parameters · 6 axioms · 0 invented entities

The central claim rests on the resummed free Hamiltonian and its exact linear modes, which are imported from Ref. [1] and recast as single integrals; no free parameters are fitted and no new physical entities are postulated.

axioms (6)
  • domain assumption The in-in formalism applies with H_int = λ_2 a³ (dot π_I^c)³; apparent extra vertices from the Legendre transform cancel at cubic order.
    Used throughout; the cancellation is cited to Ref. [9], not demonstrated in the paper (Section 'Effective field theory').
  • domain assumption Bunch–Davies initial conditions fix the Bogoliubov coefficients A^α_a and the diagonalisation condition (10).
    App. A; standard but unproved within the paper.
  • domain assumption The exact linear solutions of the coupled π–σ system from Ref. [1] are correct and are faithfully recast by the integral representation (8)–(9).
    App. A; the paper says 'We have verified...' but does not derive them or provide independent code/data.
  • domain assumption The theory is scale-invariant, with unit sound speed and M_2 = 0.
    Stated simplifications in 'Effective field theory' and 'Regime of validity'; restrict the scope of all results.
  • standard math The Bunch–Davies rotation of the in-in time integral is allowed and the leg-kernel integrals converge.
    App. B, Eq. (23); the text states 'the integral is always convergent' without proof.
  • standard math The asymptotic expansions of W^n_a for small β in Eq. (15) are valid.
    Used to derive the squeezed limit; relies on the large-u tail of the 2F1 integrand and Laplace transforms.

pith-pipeline@v1.3.0-alltime-deepseek · 11752 in / 16304 out tokens · 133501 ms · 2026-08-01T23:43:38.743689+00:00 · methodology

0 comments
read the original abstract

Using the effective field theory of multiple inflationary fluctuations, we present the first analytical calculation of the primordial bispectrum in which the quadratic mixing between curvature and isocurvature fluctuations is treated non-perturbatively. Building upon the operator representation of the exact linear solutions proposed in Ref.~\cite{Huenupi:2026abj}, we derive a simpler integral representation for these mixed mode functions. We prove that all scale-invariant tree-level bispectra reduce to a single vertex diagram, which can be evaluated with a Schwinger-parameter integral over independent pre-computable leg kernels. We showcase the power of our approach by considering the cubic time-derivative interaction $\dot{\pic}^3$, which leads to a purely single-field, equilateral phenomenology at small mixing. On the contrary, at strong mixing the obtained bispectrum shapes decorrelate from the equilateral template and become genuinely multifield, with a large amplitude, motivating a dedicated data analysis. The squeezed limit is obtained analytically in a closed form at any dimensionless mixing strength $\la$ for an isocurvature field of bare mass $m$ and features a cosmological collider signal set by $\nu_{\rm eff} = i \muf=\sqrt{9/4-m^2/H^2-\la^2}$, with an effective mass dressed by $\la$, as previously evidenced in numerical or semi-analytical calculations. Our results encompass the $\la \ll 1$ limit of usual perturbative calculations, where the amplitude of the signal is necessarily small, but they also surpass them, thus opening a new analytical window into large multifield primordial non-Gaussianities.

Figures

Figures reproduced from arXiv: 2607.15251 by Lucas Pinol.

Figure 1
Figure 1. Figure 1: Left: contribution of the contact ˙πc 3 diagram to the bispectrum shape S(k1, k2, k3), normalized to unity at the equi￾lateral configuration, for (λ, µeff ) = (2.5, 2.5). Collider oscillations of frequency µeff invade the mildly squeezed configurations with an amplitude parametrically enhanced compared to the equilateral value (which is suppressed for this parameter choice). Right: cosine correlation (14) … view at source ↗
Figure 2
Figure 2. Figure 2: Left: cosmological collider signal on isosceles triangles (k2 = k3, κ = k1/k3) at µeff = 2, in units of N and rescaled by κ −1/2 , for λ = 0.5, 2, 4. Discrete points are drawn from the exact shape (13), solid lines follow from the closed-form squeezed limit (17); dashed lines are their perturbative, λ ≪ 1 limit, Eq. (18). Right: physical complex clock amplitude rescaled by λ −2 , R −3/2Ae iδ/λ2 , as parame… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. geoSCET: Soft Theorems from Power Counting

    hep-th 2026-07 accept novelty 8.0

    geoSCET derives geometric soft theorems for scalar field theories directly from effective-field-theory power counting and proves they are exact to all orders in perturbation theory when no potential is present.

  2. Dissecting the Scalar Cosmological Collider with the Cosmic Microwave Background

    astro-ph.CO 2026-07 accept novelty 7.0

    A joint Bayesian fit of the multi-field inflationary Lagrangian to Planck bispectrum data finds no cosmological collider signal (max Δχ²=5.3) and shows weak-mixing template searches are invalid except at the smallest masses.

  3. Pushing the Primordial Frontier: Cosmological Collider Signatures at Strong Mixing

    astro-ph.CO 2026-07 conditional novelty 7.0

    Exact analytic squeezed-limit bispectra for strongly mixed two-field inflation, nonperturbative in the curvature-isocurvature mixing λ.

Reference graph

Works this paper leans on

42 extracted references · 37 linked inside Pith · cited by 3 Pith papers

  1. [1]

    Huenupi, C

    J. Huenupi, C. Mu˜ noz, G. A. Palma, and S. Sypsas, (2026), arXiv:2606.18248 [astro-ph.CO]

    Pith/arXiv arXiv 2026

  2. [2]

    Chen and Y

    X. Chen and Y. Wang, Phys. Rev. D81, 063511 (2010), arXiv:0909.0496 [astro-ph.CO]

    Pith/arXiv arXiv 2010

  3. [3]

    Chen and Y

    X. Chen and Y. Wang, JCAP04, 027 (2010), arXiv:0911.3380 [hep-th]

    Pith/arXiv arXiv 2010

  4. [4]

    Noumi, M

    T. Noumi, M. Yamaguchi, and D. Yokoyama, JHEP06, 051 (2013), arXiv:1211.1624 [hep-th]

    Pith/arXiv arXiv 2013

  5. [5]

    Arkani-Hamed and J

    N. Arkani-Hamed and J. Maldacena, (2015), arXiv:1503.08043 [hep-th]

    Pith/arXiv arXiv 2015

  6. [6]

    H. Lee, D. Baumann, and G. L. Pimentel, JHEP12, 040 (2016), arXiv:1607.03735 [hep-th]

    Pith/arXiv arXiv 2016

  7. [7]

    Arkani-Hamed, D

    N. Arkani-Hamed, D. Baumann, H. Lee, and G. L. Pi- mentel, JHEP04, 105 (2020), arXiv:1811.00024 [hep-th]

    Pith/arXiv arXiv 2020

  8. [8]

    Ach´ ucarro, J.-O

    A. Ach´ ucarro, J.-O. Gong, S. Hardeman, G. A. Palma, and S. P. Patil, Phys. Rev. D84, 043502 (2011), arXiv:1005.3848 [hep-th]

    Pith/arXiv arXiv 2011

  9. [9]

    X. Chen, Y. Wang, and Z.-Z. Xianyu, JCAP12, 006 (2017), arXiv:1703.10166 [hep-th]

    Pith/arXiv arXiv 2017

  10. [10]

    Qin and Z.-Z

    Z. Qin and Z.-Z. Xianyu, JHEP07, 001 (2023), arXiv:2301.07047 [hep-th]

    Pith/arXiv arXiv 2023

  11. [11]

    Pinol, S

    L. Pinol, S. Aoki, S. Renaux-Petel, and M. Yamaguchi, Phys. Rev. D107, L021301 (2023), arXiv:2112.05710 [hep-th]

    Pith/arXiv arXiv 2023

  12. [12]

    S. Aoki, L. Pinol, F. Sano, M. Yamaguchi, and Y. Zhu, JHEP09, 176 (2024), arXiv:2404.09547 [hep-th]

    Pith/arXiv arXiv 2024

  13. [13]

    H. An, M. McAneny, A. K. Ridgway, and M. B. Wise, JHEP06, 105 (2018), arXiv:1706.09971 [hep-ph]

    Pith/arXiv arXiv 2018

  14. [14]

    Werth, L

    D. Werth, L. Pinol, and S. Renaux-Petel, Phys. Rev. Lett.133, 141002 (2024), arXiv:2302.00655 [hep-th]

    Pith/arXiv arXiv 2024

  15. [15]

    Jazayeri, S

    S. Jazayeri, S. Renaux-Petel, and D. Werth, JCAP12, 035 (2023), arXiv:2307.01751 [hep-th]

    Pith/arXiv arXiv 2023

  16. [16]

    Pinol, S

    L. Pinol, S. Renaux-Petel, and D. Werth, JCAP02, 019 (2025), arXiv:2312.06559 [astro-ph.CO]

    Pith/arXiv arXiv 2025

  17. [17]

    Pinol, work in prep

    L. Pinol, work in prep

  18. [18]

    Creminelli, M

    P. Creminelli, M. A. Luty, A. Nicolis, and L. Senatore, JHEP12, 080 (2006), arXiv:hep-th/0606090

    Pith/arXiv arXiv 2006

  19. [19]

    Cheung, P

    C. Cheung, P. Creminelli, A. L. Fitzpatrick, J. Kaplan, and L. Senatore, JHEP03, 014 (2008), arXiv:0709.0293 [hep-th]

    Pith/arXiv arXiv 2008

  20. [20]

    Senatore and M

    L. Senatore and M. Zaldarriaga, JHEP04, 024 (2012), arXiv:1009.2093 [hep-th]

    Pith/arXiv arXiv 2012

  21. [21]

    Pinol, Phys

    L. Pinol, Phys. Rev. D110, L041302 (2024), arXiv:2405.02190 [astro-ph.CO]

    Pith/arXiv arXiv 2024

  22. [22]

    Chen and Y

    X. Chen and Y. Wang, JCAP09, 021 (2012), arXiv:1205.0160 [hep-th]

    Pith/arXiv arXiv 2012

  23. [23]

    Werth, JHEP12, 017 (2024), arXiv:2409.02072 [hep- th]

    D. Werth, JHEP12, 017 (2024), arXiv:2409.02072 [hep- th]

    Pith/arXiv arXiv 2024

  24. [24]

    Belrhali, A

    N. Belrhali, A. Poisson, S. Renaux-Petel, and D. Werth, (2026), arXiv:2604.15251 [hep-th]

    Pith/arXiv arXiv 2026

  25. [25]

    Belrhali, A

    N. Belrhali, A. Poisson, and S. Renaux-Petel, (2026), arXiv:2606.27309 [hep-th]

    Pith/arXiv arXiv 2026

  26. [26]

    Babich, P

    D. Babich, P. Creminelli, and M. Zaldarriaga, JCAP08, 009 (2004), arXiv:astro-ph/0405356

    Pith/arXiv arXiv 2004

  27. [27]

    Bodas, S

    A. Bodas, S. Kumar, and R. Sundrum, JHEP02, 079 (2021), arXiv:2010.04727 [hep-ph]. 7

    Pith/arXiv arXiv 2021

  28. [28]

    C. M. Sou, X. Tong, and Y. Wang, JHEP06, 129 (2021), arXiv:2104.08772 [hep-th]

    Pith/arXiv arXiv 2021

  29. [29]

    Senatore, K

    L. Senatore, K. M. Smith, and M. Zaldarriaga, JCAP 01, 028 (2010), arXiv:0905.3746 [astro-ph.CO]

    Pith/arXiv arXiv 2010

  30. [30]

    Akramiet al.(Planck), Astron

    Y. Akramiet al.(Planck), Astron. Astrophys.641, A9 (2020), arXiv:1905.05697 [astro-ph.CO]

    Pith/arXiv arXiv 2020

  31. [31]

    P. D. Meerburg, M. M¨ unchmeyer, J. B. Mu˜ noz, and X. Chen, JCAP03, 050 (2017), arXiv:1610.06559 [astro- ph.CO]

    Pith/arXiv arXiv 2017

  32. [32]

    Sohn, D.-G

    W. Sohn, D.-G. Wang, J. R. Fergusson, and E. P. S. Shellard, JCAP09, 016 (2024), arXiv:2404.07203 [astro- ph.CO]

    Pith/arXiv arXiv 2024

  33. [33]

    O. H. E. Philcox, K. Zhong, and S. S. Sirletti, (2025), arXiv:2511.19179 [astro-ph.CO]

    arXiv 2025

  34. [34]

    Suman, D.-G

    P. Suman, D.-G. Wang, W. Sohn, J. R. Fergusson, and E. P. S. Shellard, (2025), arXiv:2511.17500 [astro- ph.CO]

    arXiv 2025

  35. [35]

    Kumar, Q

    S. Kumar, Q. Lu, Z.-Z. Xianyu, and Y. Zhang, (2026), arXiv:2604.07434 [hep-ph]

    Pith/arXiv arXiv 2026

  36. [36]

    Garcia-Saenz, S

    S. Garcia-Saenz, S. Renaux-Petel, and J. Ronayne, JCAP07, 057 (2018), arXiv:1804.11279 [astro-ph.CO]

    Pith/arXiv arXiv 2018

  37. [37]

    Garcia-Saenz and S

    S. Garcia-Saenz and S. Renaux-Petel, JCAP11, 005 (2018), arXiv:1805.12563 [hep-th]

    Pith/arXiv arXiv 2018

  38. [38]

    A. R. Brown, Phys. Rev. Lett.121, 251601 (2018), arXiv:1705.03023 [hep-th]

    Pith/arXiv arXiv 2018

  39. [39]

    S. Aoki, D. Roest, and D. Werth, (2026), arXiv:2604.01035 [astro-ph.CO]

    arXiv 2026

  40. [40]

    McCulloch, E

    C. McCulloch, E. Pajer, and X. Tong, JHEP05, 262 (2024), arXiv:2401.11009 [hep-th]

    Pith/arXiv arXiv 2024

  41. [41]

    Bordin, P

    L. Bordin, P. Creminelli, A. Khmelnitsky, and L. Sena- tore, JCAP10, 013 (2018), arXiv:1806.10587 [hep-th]

    Pith/arXiv arXiv 2018

  42. [42]

    exchange

    J. Huenupi, C. Mu˜ noz, G. A. Palma, and S. Sypsas, work in prep. SUPPLEMENT AL MA TERIAL Appendix A: Integral representation of the ex- act linear solutions.Each field is expanded on two oscillators,X ⃗k(τ) =X α(τ, k)ˆa⃗k α + h.c.withα= 1,2 and X∈ {πc, σ}, and each mode function on the two Bunch– Davies carriers of its decoupled dynamics, e.g. πc α(τ, k)...