{"id":"3bba9df5-9884-46e3-ab70-876b753b26f8","arxiv_id":"2412.02519","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A saturating scalar self-interaction lets dark matter halos form Thomas-Fermi solitons, fuzzy solitons, or transitions between them, with even subdominant interactions seeding fuzzy solitons.","lead":"This paper simulates galaxies made of an ultralight scalar-field dark matter whose repulsive self-interactions saturate at high density. It finds that halos can form two different kinds of soliton cores and can switch from one type to the other, and that even weak self-interactions can speed up core formation.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'critical role' seeding claim rests on one random-phase realization per parameter set; with O(1) initial density fluctuations, the t~100 versus t~3000 difference could be stochastic.","rationale":"The reader's weakest_assumption identifies exactly the concern I find most load-bearing: the seeding claim is inferred from single random-phase realizations per parameter set. This is not an attack on the authors, and I do not see an internal inconsistency in the equations. The TF-to-FDM transition itself has independent support from the Gaussian ansatz of Sec. III C, which reproduces the presence of two energy minima and the disappearance of the low-density minimum near the observed transition density. The seeding claim, however, is qualitatively different: it depends on a small early density boost acting in a system with O(1) stochastic density fluctuations, and the paper provides no way to separate the effect from realization noise. No grid resolution, box size, or convergence test is given in Sec. IV B, which compounds the uncertainty for the long-time runs, although the primary issue remains the lack of ensemble statistics. Because the claim is plausible and the analytic picture is coherent, rejection would be too strong; a conditional verdict requiring ensemble checks and numerical convergence documentation is appropriate. My recommended final verdict is therefore unchanged from the reader's CONDITIONAL verdict, and I would ask for the ensemble study described in concrete_test before treating the seeding mechanism as established.","tokens_in":24291,"tokens_out":5818,"duration_ms":65123,"concrete_test":"Run at least 10 independent phase realizations for each of three cases: RTF=0.5 with rho_c=0.5, RTF=0.1 with rho_c=0.5, and a pure FDM control with lambda=0, all with the same epsilon, mass, and box parameters. For each realization, record the first time the central density exceeds a fixed threshold, e.g. 10^3, and the mass of the resulting FDM soliton. If the distributions of formation times for RTF=0.5 and RTF=0.1 overlap substantially, or if the difference is comparable to the spread within either set, the critical-role claim is not established and the paper should report the seeding result as tentative.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract and Sec. VII claim that even subdominant self-interactions can play a critical role by boosting the seed that later forms an FDM soliton. The numerical evidence for this claim is the comparison of Fig. 3 (RTF=0.5, rho_c=0.5, FDM soliton at t~100) with Fig. 9 (RTF=0.1, rho_c=0.5, FDM soliton at t~3000). Both runs use initial conditions built from Eq. (50) with random phases, and only one realization is shown per parameter set. The initial density fluctuations satisfy Eq. (53), i.e. they are of order the mean density on the de Broglie scale, so the time at which a rare peak grows into a soliton can fluctuate strongly between realizations. No ensemble is provided, and the quoted condensation time tgr~3300 in Eq. (56) is a mean kinetic-theory estimate for a pure FDM system, not a distribution. Therefore the difference in formation time could, in principle, be a large stochastic fluctuation rather than an effect of the self-interactions. The Gaussian ansatz supports the existence of TF and FDM branches and the transition between them, but it does not by itself establish the seeding mechanism, which relies on a small early density boost whose magnitude and variance are not quantified. This is the most load-bearing weakness because the seeding claim is one of the two headline results and is stated in the abstract as a general finding.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the formation and evolution of solitons in a scalar-field dark matter model with a repulsive self-interaction that saturates at high density, motivated by axion monodromy potentials. The authors use a simplified piecewise potential (15) and simulate the Gross-Pitaevskii–Poisson system in the semi-classical regime (epsilon = 0.01) for two values of the Thomas-Fermi radius (RTF = 0.5 and 0.1) and several saturation densities rho_c. They identify three regimes: a stable Thomas-Fermi soliton when rho_c is large, a directly formed FDM soliton when rho_c is small, and a TF-to-FDM transition when the soliton density exceeds rho_c. A Gaussian energy ansatz reproduces the existence and approximate locations of the TF and FDM energy minima and the disappearance of the TF minimum at a critical mass. The paper further claims that even always-subdominant self-interactions can accelerate FDM soliton formation by boosting early density fluctuations, based on the comparison of Fig. 3 (RTF = 0.5, rho_c = 0.5) with Fig. 9 (RTF = 0.1, rho_c = 0.5).","tokens_in":24683,"tokens_out":9566,"duration_ms":97335,"significance":"The paper is clearly written and provides a useful contribution to the phenomenology of scalar-field dark matter with saturating self-interactions. Its main strengths are the explicit mapping from a monodromy-type potential to a tractable nonrelativistic model, the systematic exploration of the (RTF, rho_c) parameter space, and the Gaussian ansatz, which gives a simple analytic account of the TF and FDM branches and of the transition between them. The TF-to-FDM transition at intermediate rho_c is well supported: the disappearance of the low-density energy minimum in Fig. 7 (M ~ 0.45, rho_0 ~ 5) coincides with the simulated collapse near rho_max ~ 6 in Fig. 4. The paper is also honest about its assumptions, such as neglecting the Bessel oscillations after Eq. (15) and the energy-matching ansatz in Sec. V B 3. However, the second headline claim concerning the 'critical role' of subdominant self-interactions rests on a single realization per parameter set, and the numerical simulations lack convergence tests and error estimates.","major_comments":[{"comment":"The claim that subdominant self-interactions play a critical role in seeding a later FDM soliton rests on comparing one realization with RTF = 0.5, rho_c = 0.5 (soliton at t ~ 100) with one realization with RTF = 0.1, rho_c = 0.5 (soliton at t ~ 3000). The initial conditions have order-unity density fluctuations on the de Broglie scale, as stated in Eq. (53), so the time at which a rare density peak grows into a soliton is itself a stochastic quantity. The fact that all runs share the same random-phase realization (Sec. IV A) controls for differences between parameter sets, but it does not establish that the observed boost is typical; a single common realization could be one where the effect is unusually strong or weak. The quoted gravitational condensation time t_gr ~ 3300 in Eq. (56) is a mean kinetic-theory estimate, not a measure of the realization-to-realization scatter. I recommend running an ensemble of realizations for these two parameter sets and reporting the distribution (or at least the mean and standard deviation) of FDM soliton formation times and masses, so that the reader can assess whether the t ~ 100 versus t ~ 3000 difference exceeds the stochastic scatter.","section":"Secs. V A 2, VI A 2, Figs. 3 and 9, Eqs. (50)-(53), (56)"},{"comment":"No numerical resolution or convergence information is provided. The manuscript does not state the grid size, box size, time step, or the number of eigenmodes used in the initial conditions of Eq. (48), nor does it report tests with different resolutions. This matters because the FDM soliton radius is of order epsilon^2/M ~ 0.01 (Eq. (44)) for the masses found in the simulations, so the spatial grid must resolve scales comparable to epsilon = 0.01, while density contrasts reach values of order 10^5. Without convergence tests, the quantitative values used for validation—for example the transition time t ~ 7 in Fig. 4, the final MFDM values, and the comparison with the Gaussian-ansatz critical mass—could depend on resolution. Please provide the numerical parameters and at least one convergence test for each regime (TF-stable, TF-to-FDM transition, and FDM-seeded).","section":"Sec. IV B and all simulation results"},{"comment":"The predicted final mass of the FDM soliton after the TF collapse relies on the explicit assumption Etot,FDM ~ Etot,TF, with the leftover TF matter assigned zero energy. This assumption is acknowledged in the text, but it is load-bearing for the claim that the Gaussian ansatz quantitatively explains the mass partition between the TF and FDM solitons (the 10% versus 60% cases). The agreement is only to within a factor of about 1.6 (MFDM ~ 0.08 predicted versus ~ 0.05 simulated in Sec. V B 3), so the test is not very stringent. I ask the authors to test this assumption directly from the simulation data, for instance by measuring the energy of the material that ends up in the FDM soliton and the energy of the ejected material at the transition time (t ~ 8 in Fig. 4), or to present the energy-matching estimate more cautiously as a heuristic with an explicit statement of its uncertainty.","section":"Secs. V B 3 and VI B 3, Figs. 7 and 13"}],"minor_comments":[{"comment":"The paper explicitly neglects the decaying oscillations of the Bessel function in Eq. (11) when introducing the simplified potential (15). Since this is the main modeling step connecting the results to the axion monodromy potential, a brief discussion of how the oscillations could affect the transition threshold, or one test run with the full Phi_I, would strengthen the robustness of the conclusions.","section":"Sec. II B, Eq. (15)"},{"comment":"The text states that 'the self-interactions still prevent the existence of a FDM soliton for M < 0.012' but does not explain how this threshold is obtained from the Gaussian-ansatz energy curves. Please clarify the criterion used to define existence of a soliton branch in these figures.","section":"Sec. V B 2"},{"comment":"The estimate t_gr ~ 3300 uses rho ~ 0.5 and v ~ 1, but the provenance of these values is not given. Please state explicitly how these values follow from the initial profile (46)-(47).","section":"Eq. (56)"},{"comment":"There are several typographical and formatting errors, for example 'wich' in Sec. V B 1 and some broken equation layouts in the text. A careful proofread would improve readability.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the journal's scope and the transition result is convincing. The main risk is that the seeding claim, which is highlighted in the abstract, rests on a single realization per parameter set; I would encourage the editor to ask for ensemble simulations or, at minimum, a clear statement that the effect is demonstrated for the specific realization shown rather than as a universal feature. The absence of convergence tests is also a fixable but important shortcoming."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — read this one for the saturating self-interaction. The paper studies a non-polynomial axion-monodromy-like potential that is quartic at low density but saturates above rho_c, and shows numerically and with a Gaussian ansatz that a halo can form either a TF soliton or an FDM soliton, and that a TF soliton can collapse into an FDM soliton when it exceeds the saturation density. That transition is the strongest part: the collapse at t~7 in Fig 4 matches the energy analysis (critical mass ~0.45, density ~5) and the two cases (RTF=0.5 vs 0.1) give different mass capture fractions that the energy argument explains. The Gaussian ansatz is a genuinely useful tool here; it is not fitted to the simulations, and the agreement is qualitative but credible.\n\nThe soft spot is the second headline claim: that even when self-interactions are always subdominant they play a critical role by seeding the later FDM soliton. The evidence is a comparison of two runs, one per parameter set (Fig 3 vs Fig 9). Initial density fluctuations are O(1) on the de Broglie scale (Eq 53), so the time at which a random peak grows into a soliton could easily vary by more than the factor of 30 reported. The paper does not provide an ensemble, convergence tests, or a variance estimate for the condensation time. The Gaussian ansatz supports the existence of the TF branch and the early boost, but it does not quantify the magnitude or variance of the seeding effect. So I would treat that specific claim as suggestive, not established. The other results — the TF-to-FDM transition, the energy-barrier explanation, and the caution about the hydrodynamical picture — do hold up.\n\nThe citation pattern is fine; the method follows their prior work and standard references. The simplified potential (15) drops the Bessel oscillations, which is acceptable at this level.\n\nWho is this for? People working on soliton cores in scalar-field DM, especially axion-monodromy models. It is a within-subfield advance, not a breakthrough. It deserves a serious referee: the science is sound in its core, and the seeding claim can be fixed by adding a few ensembles or softening the language. If I were refereeing, I would ask for that before acceptance.","headline":"A solid numerical study of soliton formation with a saturating self-interaction; the TF-to-FDM transition is well supported, but the 'subdominant seeding' headline is under-evidenced by single-realization runs.","tokens_in":25129,"tokens_out":2209,"would_cite":true,"duration_ms":23233,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper shows that scalar-field dark matter halos with a saturating self-interaction can form Thomas-Fermi solitons, fuzzy dark matter solitons, and an abrupt transition between the two as the central density crosses the saturation…","keywords":["scalar-field dark matter","solitons","Thomas-Fermi regime","fuzzy dark matter","self-interactions","axion monodromy","Schrödinger-Poisson system","gravitational condensation"],"falsifier":"Re-run the $R_{\\rm TF}=0.5$, $\\rho_c=0.5$ and $R_{\\rm TF}=0.1$, $\\rho_c=0.5$ simulations with many independent random-phase initial conditions and compare the distributions of fuzzy soliton formation times; if the distributions overlap substantially, the seeding role attributed to the weak self-interaction is not established.","tokens_in":24149,"feed_emoji":"🌌","tokens_out":11431,"duration_ms":97327,"temperature":0.7,"pith_summary":"The paper studies scalar-field dark matter halos whose repulsive self-interaction rises linearly with density at low density but saturates once the density exceeds a threshold $\\rho_c$, as in axion-monodromy inspired potentials. It argues that the same halo can settle into a Thomas-Fermi soliton, where self-interaction pressure balances gravity, or a fuzzy dark matter soliton, where quantum pressure balances gravity, and that when the central density crosses $\\rho_c$ the Thomas-Fermi soliton becomes unstable and collapses into the fuzzy one. It also argues that self-interactions too weak to support any soliton can still play a critical role: they give the central region a small early density boost that seeds a fuzzy soliton much earlier than it would form without them. These behaviours are captured by a simple Gaussian ansatz, which reduces the soliton problem to finding minima of an energy curve as a function of central density. This matters because the same particle model could then produce different cores in different halos, altering predictions for galactic rotation curves and small-scale structure formation.","feed_headline":"Dark-matter cores flip type as self-interactions saturate","feed_subtitle":"One density threshold decides whether a halo core forms a Thomas-Fermi soliton or a fuzzy dark matter soliton.","key_machinery":"The central objects are the dimensionless saturating self-interaction potential $\\Phi_I(\\rho) = \\lambda\\rho$ below the threshold density $\\rho_c$ and $\\Phi_I(\\rho)=\\lambda\\rho_c$ above it, together with the Gaussian ansatz $\\rho(r)=\\rho_0 e^{-r^2/R^2}$ inserted into the total energy functional at fixed mass. The ansatz turns the search for solitons into a one-dimensional minimization problem: local minima of $E_{\\rm tot}(\\rho_0)$ correspond to the two soliton types, the depth and presence of those minima tell whether a transition happens, and the energy gap between them determines how much mass falls into the new soliton.","core_discovery":"Starting from a virialized, wave-like halo in the semiclassical regime, the formation and fate of the central soliton is controlled by the relation between its central density and the saturation density $\\rho_c$ of the self-interaction. If the central density stays below $\\rho_c$, a Thomas-Fermi soliton of fixed radius $R_{\\rm TF}$ forms and is stable. If the self-interactions saturate before they can support such a soliton, the halo eventually condenses into a fuzzy dark matter soliton supported by quantum pressure. If a Thomas-Fermi soliton grows until its central density reaches $\\rho_c$, the self-interaction pressure vanishes and the soliton collapses into a much smaller, denser fuzzy soliton. The paper further claims that even when self-interactions are always subdominant, they can seed the later fuzzy soliton by boosting central density contrasts at early times, and that all of these regimes are predictable from the energy landscape of a Gaussian density profile.","pith_inferences":["A direct testable extension is to repeat the $\\rho_c=0.5$, $R_{\\rm TF}=0.5$ run with an ensemble of independent random-phase realizations; if the early boost is causal, fuzzy soliton formation times should cluster near $t\\sim100$ rather than spread toward the no-interaction condensation time $t\\sim3000$.","The same two-minimum energy structure should persist for smooth versions of the saturating potential, so the Thomas-Fermi to fuzzy collapse is likely a generic feature of saturating self-interactions rather than an artifact of the piecewise-linear model.","Astrophysically, the seeding effect implies that a self-interaction too weak to alter halo-averaged dynamics could still leave an observable imprint through the earlier appearance of compact central cores; observations tied to the presence or timing of a core could constrain the saturation density more tightly than hydrostatic arguments alone."],"forward_implications":["A single scalar-field model can produce halos whose central cores are of different types, Thomas-Fermi or fuzzy, depending on the halo's mass and formation history rather than one universal soliton profile.","When the Thomas-Fermi soliton becomes unstable, its collapse into a fuzzy soliton is fast (around $t=7$ for $R_{\\rm TF}=0.5$, $\\rho_c=3$) and ejects most of the original core mass when the two energy levels are far apart.","Self-interactions that never dominate the dynamics can still act as a catalyst, shortening fuzzy soliton formation from $t\\sim3000$ to $t\\sim100$ for the $\\rho_c=0.5$ configurations studied.","The Gaussian energy analysis predicts for each parameter set a critical soliton mass above which the Thomas-Fermi minimum disappears, so the transition threshold can be computed before running a simulation."],"supporting_citations":[{"why":"It provides the quartic self-interaction baseline and the halo initial conditions that the present simulations reproduce when the density threshold is large.","marker":"[44]"},{"why":"It supplies the gravitational condensation time (about 3300 in the dimensionless units used here) that identifies the no-boost case as pure fuzzy dark matter dynamics.","marker":"[45]"},{"why":"It gives the Thomas-Fermi soliton profile and radius scalings used for the low-density solution.","marker":"[23]"},{"why":"It is the related study of polynomial potentials with attractive and repulsive terms and provides the Gaussian ansatz approach.","marker":"[24]"},{"why":"It introduces the Gaussian ansatz for the radial profile whose energy functional underlies the analytical predictions.","marker":"[37]"},{"why":"It derives the nonrelativistic Schrödinger-Poisson equation and the soliton ground-state formalism used throughout.","marker":"[33]"},{"why":"It motivates constructing the initial wave function as a sum over eigenmodes with random phases in the semiclassical limit.","marker":"[38, 39]"}],"fun_headline_variants":["Dark matter solitons switch type at a density threshold","Self-interaction saturation flips dark matter core type","Which soliton forms? One density threshold decides","Soliton transition driven by saturation density"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The numerical evidence that always-subdominant self-interactions seed the later fuzzy soliton compares a single random-phase realization per parameter set, so the difference in formation times could in principle be a large stochastic fluctuation.","fun_headline_variants_meta":{"raw":{"variants":["Dark matter solitons switch type at a density threshold","Self-interaction saturation flips dark matter core type","Which soliton forms? One density threshold decides","Soliton transition driven by saturation density"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000456,"raw_usage":{"total_tokens":2311,"prompt_tokens":991,"completion_tokens":1320,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":607,"completion_tokens_details":{"reasoning_tokens":1259}},"tokens_in":607,"tokens_out":1320,"duration_ms":9569,"temperature":1.0,"reasoning_tokens":1259,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:21:54.743033+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the $R_{\\rm TF}=0.5$, $\\rho_c=0.5$ and $R_{\\rm TF}=0.1$, $\\rho_c=0.5$ simulations with many independent random-phase initial conditions and compare the distributions of fuzzy soliton formation times; if the distributions overlap substantially, the seeding role attributed to the weak self-interaction is not established.","supporting_citations":[{"cited_title":"Cosmological structure formation in scalar field dark matter with repulsive self-interaction: The Incredible Shrinking Jeans Mass","cited_arxiv_id":"2106.13244","evidence_quote":"It provides the quartic self-interaction baseline and the halo initial conditions that the present simulations reproduce when the density threshold is large."}],"review_version":1}