Pith. sign in

REVIEW 2 major objections 4 minor 77 references

Adding many axion Kaluza-Klein towers to the dark dimension dilutes dark-matter decay energy by 1/N, making the simplest dark-dimension dark matter scenario viable for N≳50.

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 09:41 UTC pith:W2YSNDLO

load-bearing objection A genuinely new and mostly clean idea—axiverse towers dilute visible energy injection by 1/N—but the headline N_min=50 rests on a single symmetric fragmentation benchmark; the paper deserves peer review but needs a sensitivity analysis. the 2 major comments →

arxiv 2607.27314 v1 pith:W2YSNDLO submitted 2026-07-29 hep-ph hep-th

The Dark Dimension meets the Axiverse

Kevin Langhoff , Maria Ramos , Mario Reig This is my paper
classification hep-ph hep-th
keywords dark dimensionaxiverseKaluza-Klein towersfreeze-in dark mattertower fragmentationCMB energy injectionextra dimensionsaxion dark matter
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 argues that the dark dimension—a single micron-sized extra dimension motivated by quantum-gravity consistency constraints—can be made compatible with dark matter if the universe contains a large number N of Kaluza-Klein axion towers, an 'axiverse.' It shows that while a lone graviton tower is ruled out by cosmic microwave background and warm-dark-matter bounds, the presence of N≫1 dark axion towers dilutes the energy any decaying tower injects into visible particles by a factor of 1/N. As a result, the simplest dark-dimension dark matter model becomes viable for N≳50 without hand-tuning couplings. The paper also computes freeze-in production, finding that axion and graviton towers can jointly supply the observed dark matter abundance for axion decay constant 10^12–10^16 GeV and reheating temperature 5 MeV–O(1) GeV. Along the way it develops a continuum treatment of tower fragmentation that reduces the problem to quadrature.

Core claim

The central claim is that in a dark-dimension setup with one graviton tower and N closed-string axion KK towers, the late-time energy distribution among towers is controlled by a scale-invariant attractor solution of the coupled fragmentation equations. Because all towers equilibrate, most of the dark matter mass is transferred into the N 'dark' axion towers, and any visible decays (from the graviton or the one SM-coupled axion) carry only ~1/N of the total energy. The paper quantifies this with mass fractions M_a/M_tot ∝ 1/N and M_h/M_tot ∝ N^{-(s+2)/p}, and energy-injection weights ε_a ≈ 0.69/N, ε_h ≈ 0.46/N, which directly weaken the CMB energy-injection bound. It then combines these weig

What carries the argument

The mechanism is the multi-tower fragmentation attractor: a scale-invariant solution Y_i(μ,τ)=μ^{-2}Ψ_i(μ^p τ) of the coupled Boltzmann equations for a graviton tower and N axion towers, with a single fragmentation kernel g(x) (taken as g(x)=2 for the numerical results) and decay rates Γ_i(m)=Λ_i m^p/m_*^{p-1}. Mass conservation plus 1→2 kinematics force the kernel to satisfy ∫ x g(x) dx = 1 and g(x)=g(1−x). The attractor fixes each tower's late-time mass fraction; the third moment of the mass distribution (weighted by μ^3, as needed for decay rates to SM states) gives the energy-injection weights ε_i that enter the CMB bound. The paper converts the fragmentation system into a continuum inte

Load-bearing premise

The quantitative 1/N dilution rests on the simplifying assumptions, stated in Sec. IV.B and App. SIII, that all fragmentation channels share a single universal kernel g(x)=2, that gravitons couple equally to every axion tower, and that axion towers never decay into one another; if any of these fails, the attractor mass fractions M_h/M_tot ∝ N^{-4/7} and the quoted ε_a ≈ 0.69/N, ε_h ≈ 0.46/N suppression are not guaranteed.

What would settle it

A direct calculation: solve the coupled fragmentation equations with species-dependent kernels (e.g., g_{h→aa} and g_{a→ah} differing) or with axion-axion decays allowed, and check whether the third-moment injection weights ε_h and ε_a still scale as 1/N. An observational route: if fifth-force experiments raise the lower bound on Δm_KK to about 30 meV, the required N_min grows past 10^4, which would clash with axion counts and decay constants typical of the largest string-axiverse constructions, falsifying the scenario as stated.

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

If this is right

  • If correct, dark dimension dark matter with a single micron-scale dimension is not excluded: N≳50 towers (well within string-axiverse expectations) make the model viable without tuning couplings.
  • Freeze-in production to graviton plus axion towers can account for the full dark matter abundance for f_a in 10^12–10^16 GeV and T_RH in 5 MeV–O(1) GeV, with the axion tower typically dominating when f_a is not too large.
  • CMB energy-injection and warm-dark-matter bounds translate into a concrete lower bound on the tower count: N_min ≈ 50 today, and N≳10^4 if fifth-force or kick-velocity bounds improve by factors of ~6 and ~2.4 respectively.
  • The single-tower fragmentation attractor is generalised to coupled towers, and the continuum reduction to quadrature provides a reusable analytic tool for KK-tower cosmology beyond this specific dark-dimension setup.

Where Pith is reading between the lines

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

  • The 1/N dilution mechanism may be generic: any collection of N dark species that share energy with a visible tower through fragmentation would suppress visible decay signals, so the argument could extend beyond the axiverse to other string-motivated tower sectors.
  • The paper leaves the allowed N window bounded from above as well as below: if typical string constructions prefer N well above 10^4, freeze-in overproduces DM unless f_a is suppressed, which would tension the scenario and make the viable axiverse parameter space a finite band.
  • One could test the robustness of the 1/N scaling by solving the coupled fragmentation equations with species-dependent kernels (for instance g_{h→aa} ≠ g_{a→ah}) or allowing axion-axion decays; the present paper's single-kernel benchmark does not cover these possibilities.
  • The continuum fragmentation framework might be applied to other decaying-dark-matter towers, e.g., long-lived states in warped extra dimensions, where the same attractor could predict suppressed visible signatures.

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

2 major / 4 minor

Summary. The paper combines the dark dimension scenario with a string-theoretic axiverse. It computes freeze-in production of KK graviton and axion towers through inverse decays, obtaining DM parameter space in (f_a, T_RH) for f_a ~ 10^12–10^16 GeV and T_RH ~ 5 MeV–O(1) GeV. It then models tower fragmentation via near-mass-conserving decays, derives a scale-invariant attractor for a single tower, and extends the formalism to one graviton tower plus N axion towers with a symmetric coupling structure. The central result is that the late-time energy fraction in any visible tower, and hence energy injected into the SM, is suppressed as O(1/N); the paper derives ε_a ≈ 0.69/N and ε_h ≈ 0.46/N and uses these to obtain a minimal tower number N_min ≈ 50 for viability against CMB and warm-dark-matter bounds.

Significance. If the claims are robust, the paper offers a concrete resolution of a long-standing tension in the dark dimension dark matter program: the simplest single-tower model is excluded by decaying-DM constraints, while the axiverse extension can evade them for N ≳ 50. The paper has genuine technical strengths: the single-tower attractor is checked against an exact solution for g(x)=2 (Eq. 27); App. SII gives an analytic Laplace-transform argument for the washout of initial conditions; the freeze-in integrals in Sec. III and App. SI are explicit and plausible; and the numerical iteration scheme in App. SIII is clearly described and converged. These strengths make the derivation credible within the stated benchmark. The significance is conditional, however, because the headline 1/N dilution and N_min are demonstrated only for a single symmetric fragmentation benchmark, and the paper does not quantify how they change when that benchmark is relaxed.

major comments (2)
  1. [Sec. IV.B, Eqs. (30)–(32), and App. SIII] The quantitative claims ε_a≈0.69/N, ε_h≈0.46/N (Eq. 45) and hence N_min=50 (Eq. 47) are computed only for one benchmark: a universal fragmentation kernel g(x)=2, α=1, equal graviton coupling to all axion towers, and no axion-axion tower decays. The paper itself states that these choices are made 'for simplicity' and 'need not follow' in general. The scaling exponents in Eqs. (40)–(41) explicitly depend on the small-x exponent s of the kernel, and the coefficients must also depend on the full kernel shape and on α. No sensitivity scan, analytic bound, or argument for robustness is provided. Since the abstract generalizes to 'If there are N≫1 axion towers… diluted by a factor N', this is a load-bearing step. Please add a sensitivity analysis (e.g., vary s, g(x), α; include inter-axion decays) or restrict the abstract and conclusions to the symmetric benchmark.
  2. [Sec. IV.B, Eq. (34), and Sec. V, Eq. (47)] Part of the 1/N dilution is built into the symmetric ansatz rather than being an emergent consequence of having N towers. The matrix F in Eq. (34) and the equal-Λ assumption force every axion tower to share energy equally with the visible axion tower through mass conservation, so the visible tower automatically carries O(1/N) of the total mass. If dark axion towers have different couplings, or if the visible axion has a larger decay/coupling constant, the visible fraction could be much larger and the dilution correspondingly weaker. The paper should explicitly separate the model-dependent assumption of democratic energy partitioning from the dynamical attractor computation that realizes it in this benchmark.
minor comments (4)
  1. [Abstract and Sec. V, Eq. (47)] The value N_min=50 is presented without the caveat that it is obtained for g(x)=2, α=1. Please qualify it as a benchmark-dependent estimate in the abstract and conclusions.
  2. [Reference list, [63]] Reference [63] is an acknowledgment ('We thank Miguel Montero for raising this point') rather than a citation. It should be moved to a footnote or the acknowledgments section.
  3. [Sec. V, Eq. (44)] The recasting of the CMB bound from Ref. [16] through the effective coupling λ is clear, but the meaning of λ as a multiplicative rescaling of the graviton coupling in Eq. (1) could be stated more explicitly at first use.
  4. [Sec. IV.B, Eq. (34)] The display of the matrix F is compressed; a reader not already familiar with the block structure may misinterpret the dimensions. A short sentence defining rows as daughters and columns as parents would help.

Circularity Check

0 steps flagged

No circular derivation: the 1/N dilution is a stated consequence of an explicit symmetric fragmentation model, not a fitted or self-referential input.

full rationale

The central quantities are derived, not fitted. The freeze-in abundance (Eqs. 8-13) follows from standard collision integrals with matrix elements computed from the couplings; no parameter is tuned to the DM abundance that is later 'predicted'. The multi-tower fragmentation attractor is obtained by solving the coupled Boltzmann equations (30)-(38) numerically, with the mass fractions (40)-(41) and injection weights epsilon_a ~ 0.69/N, epsilon_h ~ 0.46/N being outputs of the iteration rather than inputs. The equal sharing M_a ~ M_tot/N is a consequence of the paper's explicit modeling assumption that all N axion towers are identical and democratically coupled to the graviton ('for simplicity we use a single fragmentation kernel g for all channels... We assume gravitons couples equally to all axions'); this is a transparent model assumption, not a circular inversion in which the conclusion defines the model. CMB, warm-DM, and fifth-force bounds are imported from independent external references [16,17,47,48], and the paper does not invoke a uniqueness theorem from its own prior work. The only self-citation ([64], for a statistical factor in the freeze-in integral) is standard and not load-bearing. Robustness to kernel/coupling variations is a legitimate scientific concern, but it is a limitation of the benchmark, not circularity.

Axiom & Free-Parameter Ledger

4 free parameters · 6 axioms · 0 invented entities

No new particles or forces are introduced: axion KK towers and graviton KK towers are already present in the dark-dimension/axiverse literature. The paper's contributions are a new fragmentation formalism and its application to a multi-tower axiverse. The main 'free parameters' are modeling choices (p, g, α, s) rather than fitted constants; physical parameters like Δm_KK, f_a, T_RH, N are scanned, not fit.

free parameters (4)
  • p = 7/2
    Exponent in KK decay rate Γ(m)=m^p/m_*^{p-1}. Adopted from prior dark-dimension fragmentation literature [9,17]; controls attractor shape and N-scaling exponents, so the quantitative predictions depend on it.
  • fragmentation kernel g(x) = g(x)=2 (uniform)
    Chosen because it is exactly solvable and satisfies mass conservation and 1→2 symmetry; not derived from a UV string model. Numerical results for ε_a, ε_h and N_min use this kernel.
  • coupling ratio α = 1
    O(1) ratio of a^2h to h^3 decay couplings appearing in Λ_a and Λ_h; set to 1 for all numerical results, with no scan.
  • small-x exponent s = 0 (for g=2)
    Enters M_h/M_tot ∝ N^{-(s+2)/p}; set implicitly by choosing g=2. No independent derivation.
axioms (6)
  • domain assumption Dark dimension scenario: a single micron-scale extra dimension with KK mass splitting Δm_KK ∼ Λ^{1/4} ≈ meV, and a tower of KK gravitons.
    Borrowed from swampland/dark dimension literature [7-9]; the paper's phenomenology is built on this. Not independently established.
  • domain assumption String axiverse: closed-string axions (from C_p/B fields) propagate in the dark dimension and form KK towers with the same mass scaling as the graviton; only one axion couples to SM gauge bosons, N−1 are dark.
    Motivated by string compactifications [18-21,29,30]; the paper explicitly assumes this ("We assume throughout that only one axion couples to the SM").
  • domain assumption Freeze-in initial conditions: bulk is empty after reheating and does not thermalize; production is dominated by inverse decays.
    Standard UV freeze-in assumption [45]; no thermalization check for the axion parameter range.
  • domain assumption Near-perfect mass conservation in KK fragmentation (approximate translation symmetry along the dark dimension), with 1→2 decays.
    Stated in Sec. IV; needed so DM is not too warm and late-time abundance is set by redshift alone. Not derived from the UV.
  • ad hoc to paper Tower fragmentation model: universal homogeneous kernel g, equal graviton coupling to all axion towers, no axion-axion tower decays; decay rates Γ_i = Λ_i m^p/m_*^{p-1}.
    Eqs. (19), (30)-(32). Described "for simplicity"; determines the attractor and the 1/N dilution factors. The paper does not derive these from string theory.
  • domain assumption Misalignment contribution to axion-tower DM is negligible.
    Sec. III: assumed under instantaneous reheating with T_RH ~ O(1) GeV; cases where H_inf fails conditions require further study.

pith-pipeline@v1.3.0-daily-deepseek · 17091 in / 16996 out tokens · 135544 ms · 2026-08-01T09:41:31.627879+00:00 · methodology

0 comments
read the original abstract

We explore the cosmological implications of combining dark dimension scenarios with an axiverse. If gauge sectors are realized on branes, towers of Kaluza-Klein (KK) excitations of closed string axions can propagate through the dark dimension in addition to the tower of graviton excitations. This modifies cosmology in two ways. First, if any of these axion towers interact with the standard model (SM) plasma, they can significantly alter the freeze-in production of the cosmological abundance of tower states. Freeze-in to graviton and axion towers can provide all of dark matter (DM) for an axion decay constant in $10^{12} \text{ GeV }\lesssim f_a\lesssim 10^{16}\,\text{ GeV }$ and reheating temperatures $5\text{ MeV }\lesssim T_{\rm RH} \lesssim O(1)\,\text{ GeV }$. Second, different towers fragment into each other and redistribute energy; each tower's fraction of energy at late times is fixed by their interactions. If there are $N\gg1$ axion towers, the energy visibly injected into the SM by any decaying tower is diluted by a factor of $N$. This suppression offers a simple realization of how dark dimension dark matter can avoid strong cosmological constraints which rule out the simplest models. In the process of this exploration we develop a continuum approach to evaluating tower fragmentation which offers insight and aids numerical calculations by reducing the problem to quadrature.

Figures

Figures reproduced from arXiv: 2607.27314 by Kevin Langhoff, Maria Ramos, Mario Reig.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Late time constraints on dark dimension dark matter with [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Allowed region (shaded) for dark dimension dark matter in the ( [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

77 extracted references · 50 linked inside Pith

  1. [1]

    Choose a trial solution Ψ 1(ξ)

  2. [2]

    (26); normalize to conserve mass

    Fori≥1 evaluateG i(ξ) using Ψ i(ξ) and solve for Ψi+1(ξ) using eq. (26); normalize to conserve mass

  3. [3]

    reasonable

    Repeat step 2 until a threshold is reached. Results of this method are shown in Fig. S1. To test this method, we compare with the exactly solvable solution Ψuniform(ξ) = pM Γ 2 p ξ 2 p e−ξ ,forg(x) = 2,(27) where we define the conserved total mass of all modes M= Z ∞ 0 µ Y(µ, τ)dµ= 1 p Z ∞ 0 Ψ(ξ)dlogξ .(28) Initial conditions, and therefore solutions at g...

  4. [4]

    L¨ ust, E

    D. L¨ ust, E. Palti, and C. Vafa, Phys. Lett. B797, 134867 (2019), arXiv:1906.05225 [hep-th]

    Pith/arXiv arXiv 2019

  5. [5]

    Ooguri and C

    H. Ooguri and C. Vafa, Nucl. Phys. B766, 21 (2007), arXiv:hep-th/0605264

    Pith/arXiv arXiv 2007

  6. [6]

    S.-J. Lee, W. Lerche, and T. Weigand, JHEP02, 190 (2022), arXiv:1910.01135 [hep-th]

    Pith/arXiv arXiv 2022

  7. [7]

    S.-J. Lee, W. Lerche, and T. Weigand, JHEP02, 096 (2022), arXiv:1904.06344 [hep-th]

    Pith/arXiv arXiv 2022

  8. [8]

    Vafa, (2024), arXiv:2402.00981 [hep-ph]

    C. Vafa, (2024), arXiv:2402.00981 [hep-ph]

    Pith/arXiv arXiv 2024

  9. [9]

    Obied, H

    G. Obied, H. Ooguri, L. Spodyneiko, and C. Vafa, (2018), arXiv:1806.08362 [hep-th]

    Pith/arXiv arXiv 2018

  10. [10]

    Bedroya and C

    A. Bedroya and C. Vafa, JHEP09, 123 (2020), arXiv:1909.11063 [hep-th]

    Pith/arXiv arXiv 2020

  11. [11]

    Montero, C

    M. Montero, C. Vafa, and I. Valenzuela, JHEP02, 022 (2023), arXiv:2205.12293 [hep-th]

    Pith/arXiv arXiv 2023

  12. [12]

    R. N. Mohapatra, S. Nussinov, and A. Perez-Lorenzana, Phys. Rev. D68, 116001 (2003), arXiv:hep-ph/0308051

    Pith/arXiv arXiv 2003

  13. [13]

    Gonzalo, M

    E. Gonzalo, M. Montero, G. Obied, and C. Vafa, JHEP 11, 109 (2023), arXiv:2209.09249 [hep-ph]

    Pith/arXiv arXiv 2023

  14. [14]

    G. F. Giudice, R. Rattazzi, and J. D. Wells, Nucl. Phys. B544, 3 (1999), arXiv:hep-ph/9811291

    Pith/arXiv arXiv 1999

  15. [15]

    L. J. Hall and D. Tucker-Smith, Phys. Rev. D60, 085008 (1999), arXiv:hep-ph/9904267

    Pith/arXiv arXiv 1999

  16. [16]

    J. A. P. Law-Smith, G. Obied, A. Prabhu, and C. Vafa, JHEP06, 047 (2024), arXiv:2307.11048 [hep-ph]

    Pith/arXiv arXiv 2024

  17. [17]

    K. R. Dienes and B. Thomas, Phys. Rev. D85, 083523 (2012), arXiv:1106.4546 [hep-ph]

    Pith/arXiv arXiv 2012

  18. [18]

    K. R. Dienes and B. Thomas, Phys. Rev. D85, 083524 (2012), arXiv:1107.0721 [hep-ph]

    Pith/arXiv arXiv 2012

  19. [19]

    K. R. Dienes and B. Thomas, Phys. Rev. D86, 055013 (2012), arXiv:1203.1923 [hep-ph]

    Pith/arXiv arXiv 2012

  20. [20]

    Cicoli, M

    M. Cicoli, M. Goodsell, and A. Ringwald, JHEP10, 146 (2012), arXiv:1206.0819 [hep-th]

    Pith/arXiv arXiv 2012

  21. [21]

    Obied, C

    G. Obied, C. Dvorkin, E. Gonzalo, and C. Vafa, Phys. Rev. D109, 063540 (2024), arXiv:2311.05318 [astro- ph.CO]

    Pith/arXiv arXiv 2024

  22. [22]

    Svrcek and E

    P. Svrcek and E. Witten, JHEP06, 051 (2006), arXiv:hep-th/0605206

    Pith/arXiv arXiv 2006

  23. [23]

    Arvanitaki, S

    A. Arvanitaki, S. Dimopoulos, S. Dubovsky, N. Kaloper, and J. March-Russell, Phys. Rev. D81, 123530 (2010), arXiv:0905.4720 [hep-th]

    Pith/arXiv arXiv 2010

  24. [24]

    J. J. Heckman, C. Vafa, T. Weigand, and F. Xu, Phys. Rev. D111, 046014 (2025), arXiv:2409.01405 [hep-th]

    Pith/arXiv arXiv 2025

  25. [25]

    Gendler, D

    N. Gendler, D. J. E. Marsh, L. McAllister, and J. Moritz, 10 JCAP09, 071 (2024), arXiv:2309.13145 [hep-th]

    Pith/arXiv arXiv 2024

  26. [26]

    A. P. Braun, M. Cicoli, R. Milioli, and R. Valandro, (2026), arXiv:2606.19440 [hep-th]

    Pith/arXiv arXiv 2026

  27. [27]

    J. H. Schwarz, Int. J. Mod. Phys. A39, 2447010 (2024), arXiv:2403.12899 [hep-th]

    Pith/arXiv arXiv 2024

  28. [28]

    Choi and J

    K. Choi and J. E. Kim, Phys. Lett. B154, 393 (1985), [Erratum: Phys.Lett.B 156, 452 (1985)]

    1985

  29. [29]

    Reig and I

    M. Reig and I. Ruiz, (2025), arXiv:2510.25832 [hep-th]

    Pith/arXiv arXiv 2025

  30. [30]

    Blumenhagen and A

    R. Blumenhagen and A. Paraskevopoulou, (2026), arXiv:2605.11068 [hep-th]

    Pith/arXiv arXiv 2026

  31. [31]

    Witten, Phys

    E. Witten, Phys. Lett. B149, 351 (1984)

    1984

  32. [32]

    R. D. Peccei and H. R. Quinn, Phys. Rev. Lett.38, 1440 (1977)

    1977

  33. [33]

    Demirtas, C

    M. Demirtas, C. Long, L. McAllister, and M. Stillman, JHEP04, 138 (2020), arXiv:1808.01282 [hep-th]

    Pith/arXiv arXiv 2020

  34. [34]

    S. V. P. Fallon, J. Halverson, L. McAllister, and Y. Zhu, (2025), arXiv:2511.20458 [hep-th]

    arXiv 2025

  35. [35]

    L. E. Ibanez and A. M. Uranga,String theory and par- ticle physics: An introduction to string phenomenology (Cambridge University Press, 2012)

    2012

  36. [36]

    Craig and M

    N. Craig and M. Kongsore, Phys. Rev. D111, 015047 (2025), arXiv:2408.10295 [hep-ph]

    Pith/arXiv arXiv 2025

  37. [37]

    R. D. Peccei and H. R. Quinn, Phys. Rev. D16, 1791 (1977)

    1977

  38. [38]

    Weinberg, Phys

    S. Weinberg, Phys. Rev. Lett.40, 223 (1978)

    1978

  39. [39]

    Wilczek, Phys

    F. Wilczek, Phys. Rev. Lett.40, 279 (1978)

    1978

  40. [40]

    Holman, S

    R. Holman, S. D. Hsu, T. W. Kephart, E. W. Kolb, R. Watkins,et al., Phys.Lett.B282, 132 (1992), arXiv:hep-ph/9203206 [hep-ph]

    Pith/arXiv arXiv 1992

  41. [41]

    H. M. Georgi, L. J. Hall, and M. B. Wise, Nucl. Phys. B192, 409 (1981)

    1981

  42. [42]

    Kamionkowski and J

    M. Kamionkowski and J. March-Russell, Phys.Lett. B282, 137 (1992), arXiv:hep-th/9202003 [hep-th]

    Pith/arXiv arXiv 1992

  43. [43]

    S. M. Barr and D. Seckel, Phys. Rev. D46, 539 (1992)

    1992

  44. [44]

    Gendler and C

    N. Gendler and C. Vafa, JHEP12, 127 (2024), arXiv:2404.15414 [hep-th]

    Pith/arXiv arXiv 2024

  45. [45]

    J. N. Benabou, K. Fraser, M. Reig, and B. R. Safdi, (2025), arXiv:2505.15884 [hep-ph]

    Pith/arXiv arXiv 2025

  46. [46]

    Reece, JHEP07, 130 (2025), arXiv:2406.08543 [hep- ph]

    M. Reece, JHEP07, 130 (2025), arXiv:2406.08543 [hep- ph]

    Pith/arXiv arXiv 2025

  47. [47]

    Reece, T

    M. Reece, T. Rudelius, and C. Tudball, (2026), arXiv:2606.21650 [hep-th]

    Pith/arXiv arXiv 2026

  48. [48]

    J. G. Lee, E. G. Adelberger, T. S. Cook, S. M. Fleischer, and B. R. Heckel, Phys. Rev. Lett.124, 101101 (2020), arXiv:2002.11761 [hep-ex]

    Pith/arXiv arXiv 2020

  49. [49]

    L. J. Hall, K. Jedamzik, J. March-Russell, and S. M. West, JHEP03, 080 (2010), arXiv:0911.1120 [hep-ph]

    Pith/arXiv arXiv 2010

  50. [50]

    Aghanimet al.(Planck), Astron

    N. Aghanimet al.(Planck), Astron. Astrophys.641, A6 (2020), [Erratum: Astron.Astrophys. 652, C4 (2021)], arXiv:1807.06209 [astro-ph.CO]

    Pith/arXiv arXiv 2020

  51. [51]

    Tan, S.-Q

    W.-H. Tan, S.-Q. Yang, C.-G. Shao, J. Li, A.-B. Du, B.- F. Zhan, Q.-L. Wang, P.-S. Luo, L.-C. Tu, and J. Luo, Phys. Rev. Lett.116, 131101 (2016)

    2016

  52. [52]

    Reig, JHEP09, 207 (2021), arXiv:2104.09923 [hep- ph]

    M. Reig, JHEP09, 207 (2021), arXiv:2104.09923 [hep- ph]

    Pith/arXiv arXiv 2021

  53. [53]

    Hasegawa, N

    T. Hasegawa, N. Hiroshima, K. Kohri, R. S. L. Hansen, T. Tram, and S. Hannestad, JCAP12, 012 (2019), arXiv:1908.10189 [hep-ph]

    Pith/arXiv arXiv 2019

  54. [54]

    P. F. de Salas, M. Lattanzi, G. Mangano, G. Miele, S. Pastor, and O. Pisanti, Phys. Rev. D92, 123534 (2015), arXiv:1511.00672 [astro-ph.CO]

    Pith/arXiv arXiv 2015

  55. [55]

    P. W. Graham and A. Scherlis, Phys. Rev. D98, 035017 (2018), arXiv:1805.07362 [hep-ph]

    Pith/arXiv arXiv 2018

  56. [56]

    K. C. Y. Ng, B. M. Roach, K. Perez, J. F. Beacom, S. Ho- riuchi, R. Krivonos, and D. R. Wik, Phys. Rev. D99, 083005 (2019), arXiv:1901.01262 [astro-ph.HE]

    Pith/arXiv arXiv 2019

  57. [57]

    T. Han, J. D. Lykken, and R.-J. Zhang, Phys. Rev. D 59, 105006 (1999), arXiv:hep-ph/9811350

    Pith/arXiv arXiv 1999

  58. [58]

    Boyarsky, J

    A. Boyarsky, J. Nevalainen, and O. Ruchayskiy, Astron. Astrophys.471, 51 (2007), arXiv:astro-ph/0610961

    Pith/arXiv arXiv 2007

  59. [59]

    Boyarsky, D

    A. Boyarsky, D. Iakubovskyi, O. Ruchayskiy, and V. Savchenko, Mon. Not. Roy. Astron. Soc.387, 1361 (2008), arXiv:0709.2301 [astro-ph]

    Pith/arXiv arXiv 2008

  60. [60]

    Calore, A

    F. Calore, A. Dekker, P. D. Serpico, and T. Siegert, Mon. Not. Roy. Astron. Soc.520, 4167 (2023), [Er- ratum: Mon.Not.Roy.Astron.Soc. 538, 132 (2025)], arXiv:2209.06299 [hep-ph]

    arXiv 2023

  61. [61]

    Current bounds, assuming long-lived particles that do not get trapped within the star, constrain a single axion coupling to photons to be (1/fa)single ≲10 −5 GeV−1 [62]

    for a recent study of KK gravitons), being inde- pendent of the DM fraction, could provide constraints on visible axion towers. Current bounds, assuming long-lived particles that do not get trapped within the star, constrain a single axion coupling to photons to be (1/fa)single ≲10 −5 GeV−1 [62]. In this scenario, for typ- ical supernovae core temperature...

  62. [62]

    B. M. Roach, K. C. Y. Ng, K. Perez, J. F. Beacom, S. Ho- riuchi, R. Krivonos, and D. R. Wik, Phys. Rev. D101, 103011 (2020), arXiv:1908.09037 [astro-ph.HE]

    Pith/arXiv arXiv 2020

  63. [63]

    J. W. Foster, M. Kongsore, C. Dessert, Y. Park, N. L. Rodd, K. Cranmer, and B. R. Safdi, Phys. Rev. Lett. 127, 051101 (2021), arXiv:2102.02207 [astro-ph.CO]

    Pith/arXiv arXiv 2021

  64. [64]

    B. M. Roach, S. Rossland, K. C. Y. Ng, K. Perez, J. F. Beacom, B. W. Grefenstette, S. Horiuchi, R. Krivonos, and D. R. Wik, Phys. Rev. D107, 023009 (2023), arXiv:2207.04572 [astro-ph.HE]

    Pith/arXiv arXiv 2023

  65. [65]

    Hardy, A

    E. Hardy, A. Sokolov, and H. Stubbs, JHEP03, 029 (2026), arXiv:2510.18975 [hep-ph]

    arXiv 2026

  66. [66]

    Lucente, P

    G. Lucente, P. Carenza, T. Fischer, M. Giannotti, and A. Mirizzi, JCAP12, 008 (2020), arXiv:2008.04918 [hep- ph]

    arXiv 2020

  67. [67]

    We thank Miguel Montero for raising this point

  68. [68]

    Langhoff, N

    K. Langhoff, N. J. Outmezguine, and N. L. Rodd, Phys. Rev. Lett.129, 241101 (2022), arXiv:2209.06216 [hep- ph]

    Pith/arXiv arXiv 2022

  69. [69]

    The Dark Dimension meets the Axiverse

    A. Bazavovet al.(HotQCD), Phys. Lett. B795, 15 (2019), arXiv:1812.08235 [hep-lat]. 1 SUPPLEMENTARY MATERIAL “The Dark Dimension meets the Axiverse” Kevin Langhoff, Maria Ramos, and Mario Reig SI. CALCULA TION OF FREEZE-IN RELIC ABUNDANCE This appendix calculates the freeze-in abundance quoted in (12) and (13). A mode of massmevolves according to the Boltz...

    Pith/arXiv arXiv 2019

  70. [70]

    Distributions converge to the attractor solution

  71. [71]

    Initial conditions get washed out leaving only the information of the total mass (assuming mass conservation)

  72. [72]

    Consider uniform fragmentation of a single tower withF(m, m′) =F(m) form > m′

    The time scale to reach the attractor solution is identified with the lifetime of initial-state particles. Consider uniform fragmentation of a single tower withF(m, m′) =F(m) form > m′. We Laplace transform Eq. (16) s bY(m, s)−Y 0(m) =−Γ(m) bY(m, s) + Z ∞ m dm′ Γ(m′)F(m ′) bY(m ′, s),(S9) whereY 0(m) =Y 0(m, t= 0). We then differentiate with respect tom(d...

  73. [73]

    Initialize Ψ (1) i (ξ) =ξ (s+2)/p e−Λiξ, with the small-ξexponent fixed by mass conservation

  74. [74]

    EvaluateG[Ψ (j) h ] andG[Ψ (j) a ] and assemble the sources Eq. (S18)

  75. [75]

    Advance Eq. (S19) from Ψi(ξmin) = 0 toξ max, taking the integral over each cell in closed form: withδk =ξ k+1−ξk it equalsA k Gi(ξk)+(A k −Bk) Gi(ξk+1)−G i(ξk) , whereA k = (1−e −yk )/Λi andB k = [1−(1+y k)e−yk ]/(Λ2 i δk), the latter evaluated by its seriesδ k( 1 2 − yk 3 +. . .) foryk ≲10 −4 to avoid cancellation

  76. [76]

    Normalize both profiles by the total massM tot =M h +NM a from Eq. (28)

  77. [77]

    A few hundred passes leave the profiles stationary at the 10 −4 level, and every number quoted below is unchanged fornbetween 1500 and 6000

    Repeat from step 2 until stationary. A few hundred passes leave the profiles stationary at the 10 −4 level, and every number quoted below is unchanged fornbetween 1500 and 6000. Figure S1 validates the method on a single tower (Λ = 1), starting from a trial peaked an order of magnitude above the attractor. The sixth iterate is indistinguishable from it on...