REVIEW
Cosmology of self-interacting ultralight dark matter: simultaneous shifts in the Hubble constant and structure growth
Cosmology and Nongalactic Astrophysics · Submitted 21 Oct 2026 · 46 references
Pith's verdict: First CLASS-level demonstration that a single self-interacting ULDM field can simultaneously raise fixed-θ* H0 and suppress late-time growth, but the growth numbers rest on an interpolation the authors admit is not an exact closure.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
A single self-interacting ultralight scalar field can, through its own successive dynamical regimes, raise the fixed-acoustic-scale value of the Hubble constant while suppressing late-time structure growth.
The reading
What carries the argument
The background machinery is the three-regime evolution of the homogeneous field, with the ratio of the Hubble rate to the effective oscillation frequency picking out the regime; the combination $\tilde m \equiv m/\lambda^{1/4}$ sets the temperature of the quartic-to-quadratic transition, and the order-eV values used here put that transition close to matter-radiation equality, where a radiation-like episode most efficiently shrinks the sound-horizon integrand. The $H_0$ shift is defined by the fixed-acoustic-scale procedure: for each benchmark the Hubble constant is adjusted until $\theta_* = r_s(z_*)/D_M(z_*)$ equals its $\Lambda$CDM reference value. The growth shift is carried by the perturbation pipeline, in which the exact Klein-Gordon system runs from the physical oscillation onset at $z \simeq 1.1\times10^9$ up to a numerical handoff at $z = 10^6$, where a causal oscillation average of the scalar stress-energy $(\bar\rho_\phi, \bar p_\phi, \delta\rho_\phi, q_\phi)$ initializes an effective fluid and all other variables transfer continuously. The post-handoff pressure perturbation is advanced with the scale-dependent sound speed $c^2_{s,\phi}(k,a) = \bigl(k^2/4m^2a^2 + \tfrac34\lambda\bar\rho_\phi/m^4\bigr)/\bigl(1 + k^2/4m^2a^2 + \tfrac94\lambda\bar\rho_\phi/m^4\bigr)$, whose limits recover $c^2_{s,\phi}\to 0$ in the late quadratic low-$k$ regime and $c^2_{s,\phi}\to 1/3$ in the quartic regime, and this closure is what produces the scale-dependent suppression of $P(k)$.
Extended reading notes
Core claim
The authors' central claim is that the self-interacting potential $V(\phi) = \tfrac12 m^2\phi^2 + \tfrac{\lambda}{4}\phi^4$ with $\lambda > 0$ produces, from one field, the two complementary effects that earlier work had to build from separate components: the quartic-dominated oscillatory phase behaves as radiation with $\langle w_\phi\rangle = 1/3$, increasing the pre-recombination expansion rate and reducing the comoving sound horizon, while the subsequent quadratic phase with $\langle w_\phi\rangle = 0$ turns the same field into dark matter whose scale-dependent effective pressure suppresses late-time clustering. Evolving the background and linear perturbations in CLASS for a mixed CDM+ULDM dark sector, with the microscopic mass $m = 10^{-22}$ eV and the quartic-to-quadratic scale $\tilde m \equiv m/\lambda^{1/4} \simeq 1$ eV, the paper fixes the acoustic angular scale $\theta_*$ to its $\Lambda$CDM reference value and finds that increasing the present-day scalar fraction $f_{\rm UDM}$ from 0.05 to 0.20 raises $H_0$ from 67.9700 to 69.1022 km s$^{-1}$ Mpc$^{-1}$, reduces $r_s(z_{\rm rec})$ from 144.4907 to 144.0259 Mpc, and lowers $\sigma_8$ from 0.780514 to 0.619318. The matter power spectrum is suppressed scale-dependently — at $k = 0.1$ Mpc$^{-1}$ the ratio $P/P_{\Lambda\mathrm{CDM}}$ falls to 0.532 and at $k = 1$ Mpc$^{-1}$ to 0.289 for $f_{\rm UDM} = 0.20$ — and the CMB temperature residual stays smooth while reaching about 8.9%. Stated on the paper's own terms, the main result is the numerical realization of a single self-interacting-ULDM mechanism that links an early reduction of the sound horizon to a later suppression of structure growth, shown to survive a stress-energy-consistent treatment of linear perturbations.
Load-bearing premise
The load-bearing premise is that the effective sound-speed formula in Eq. (4.3), which the paper itself describes as a physically motivated interpolation rather than an exact closure through the quartic-to-quadratic transition, correctly transfers the scalar's perturbation pressure across that transition; the formula is already in use at the numerical handoff $z = 10^6$, well before the transition at $z_m \simeq 7.6\times10^3$–$1.2\times10^4$, and the quoted $\sigma_8$ values and the scale dependence of the matter-power suppression follow from it.
Editorial extensions
If this is right
- Because the scalar fraction $f_{\rm UDM}$ controls both shifts, a joint fit must find that the $H_0$ gain and the $\sigma_8$ loss are correlated consequences of one parameter; a data set that permits one but forbids the other would sharply constrain the model.
- The scale dependence of the matter-power suppression is a concrete prediction: at fixed $\tilde m = 1$ eV the ratio $P/P_{\Lambda\mathrm{CDM}}$ at $k = 0.1$ Mpc$^{-1}$ falls from 0.885 to 0.532 as $f_{\rm UDM}$ goes from 0.05 to 0.20, and at $k = 1$ Mpc$^{-1}$ it reaches 0.289, so the model can be probed by comparing suppression at different wavenumbers rather than by $\sigma_8$ alone.
- The CMB temperature residual grows smoothly with the fraction, to about 8.9% for $f_{\rm UDM} = 0.20$; a full TT/TE/EE likelihood, which the paper leaves for future work, will decide whether such residuals are acceptable and will thereby bound the fraction and the achievable $H_0$ shift.
- Over the tested interval the transition scale is a milder knob than the fraction: at $f_{\rm UDM} = 0.10$, moving $\tilde m$ from 0.9 to 1.1 eV shifts $H_0$ by about 0.2 km s$^{-1}$ Mpc$^{-1}$ and $\sigma_8$ by about 0.012, so the main lever on both shifts is the scalar abundance.
- The mixed dark sector is what keeps the model safe at small scales: with $f_{\rm UDM} < 1$ the CDM component preserves efficient large-scale clustering, whereas the same scalar at full dark-matter abundance would over-suppress structure.
Reading between the lines
- An implication the paper leaves implicit: the largest fraction it studies is also the most exposed, since its roughly 9% smooth TT residual is the first thing a CMB likelihood would test; if that residual is rejected, the surviving fractions would buy only about 0.1–0.4 km s$^{-1}$ Mpc$^{-1}$ of $H_0$ gain (the 0.05 and 0.10 cases), which would cap how much of the Hubble tension this mechanism can
- The quoted $\sigma_8$ values rest on the Eq. (4.3) closure being adequate across the quartic-to-quadratic transition; a direct numerical test would be to keep the exact Klein-Gordon perturbation system active through $z_m$ instead of switching at $z = 10^6$, and to check whether the suppression amplitudes and their scale dependence survive.
- Because $\lambda = (m/\tilde m)^4 \sim 10^{-88}$ is tied to the two mass scales, the same $\tilde m$ that sets the cosmological transition also sets the kpc-scale self-interaction-supported Jeans length quoted in the paper, so cosmological and galactic constraints on $\tilde m$ can be combined into a single-parameter consistency test.
- The paper states it does not impose a BBN or $\Delta N_{\rm eff}$ constraint on its benchmarks; checking the radiation-like quartic phase against light-element abundances and the effective number of relativistic species is an open consistency test that could independently bound the allowed fraction.
Original abstract
We study the cosmology of self-interacting ultralight dark matter. The scalar field evolves from an early Hubble-frozen phase to a quartic-dominated radiation-like phase and finally to a quadratic matter-like phase. The intermediate phase increases the pre-recombination expansion rate and reduces the sound horizon, while the late phase allows the same field to contribute to dark matter. We implement the background and linear perturbations of a mixed cold-dark-matter and ultralight-dark-matter cosmology in CLASS. For $m=10^{-22}\,{\rm eV}$ and $\tilde m\simeq1\,{\rm eV}$, increasing the ultralight-dark-matter fraction raises the fixed-$\theta_*$ value of $H_0$, lowers the late-time clustering amplitude, strengthens the scale-dependent matter-power suppression, and produces a smooth CMB-temperature residual reaching about $9\%$ for the largest fraction considered. The model therefore generates simultaneous background and growth shifts in the directions relevant to the Hubble and structure-growth tensions. Establishing whether these shifts improve a joint fit requires a likelihood-level observational analysis.
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
We study the cosmology of self-interacting ultralight dark matter. The scalar field evolves from an early Hubble-frozen phase to a quartic-dominated radiation-like phase and finally to a quadratic matter-like phase. The intermediate phase increases the pre-recombination expansion rate and reduces the sound horizon, while the late phase allows the same field to contribute to dark matter. We implement the background and linear perturbations of a mixed cold-dark-matter and ultralight-dark-matter cosmology in CLASS. For $m=10^{-22}\,{\rm eV}$ and $\tilde m\simeq1\,{\rm eV}$, increasing the ultralight-dark-matter fraction raises the fixed-$\theta_*$ value of $H_0$, lowers the late-time clustering amplitude, strengthens the scale-dependent matter-power suppression, and produces a smooth CMB-temperature residual reaching about $9\%$ for the largest fraction considered. The model therefore generates simultaneous background and growth shifts in the directions relevant to the Hubble and structure-growth tensions. Establishing whether these shifts improve a joint fit requires a likelihood-level observational analysis.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | We study the cosmology of self-interacting ultralight dark matter. | contribution 0.92 | 0.87 | 0.02 | asserted 0.70 |
| s1 | The scalar field evolves from an early Hubble-frozen phase to a quartic-dominated radiation-like phase and finally to a quadratic matter-like phase. | contribution 0.34 | 0.45 | 0.04 | asserted 1.00 |
| s2 | The intermediate phase increases the pre-recombination expansion rate and reduces the sound horizon, while the late phase allows the same field to contribute to dark matter. | interpretation 0.47 | 0.52 | 0.06 | asserted 1.00 |
| s3 | We implement the background and linear perturbations of a mixed cold-dark-matter and ultralight-dark-matter cosmology in CLASS. | method 0.99 | 0.94 | 0.02 | asserted 1.00 |
| s4 | For $m=10^{-22}\,{\rm eV}$ and $\tilde m\simeq1\,{\rm eV}$, increasing the ultralight-dark-matter fraction raises the fixed-$\theta_*$ value of $H_0$, lowers the late-time clustering amplitude, strengthens the scale-dependent matter-power suppression, and produces a smooth CMB-temperature residual reaching about $9\%$ for the largest fraction considered. | result 1.00 | 0.91 | 0.30 | asserted 1.00 |
| s5 | The model therefore generates simultaneous background and growth shifts in the directions relevant to the Hubble and structure-growth tensions. | interpretation 0.99 | 0.82 | 0.75 | asserted 1.00 |
| s6 | Establishing whether these shifts improve a joint fit requires a likelihood-level observational analysis. | future work 0.67 | 0.11 | 0.09 | unclear 0.52 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
Figures
Figures reproduced from arXiv:2610.29582 by the authors.
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.