{"id":"b00a6154-7d43-419e-9038-5b7a0fbf63d6","arxiv_id":"2504.16910","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"This paper constructs a GeV-scale self-interacting dark matter model with late scalar decays that simultaneously sets the relic abundance and yields an observable Delta Neff in reach of CMB-S4.","lead":"A model-building paper proposes a GeV-scale self-interacting dark matter particle in a new dark gauge sector, whose dense cores could explain small galaxy anomalies. The same setup produces extra dark radiation that future cosmic microwave background experiments could detect.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"As written, Eq. (1) violates the unbroken U(1)_D: with Qχ=Qφ and QνS=0, the operator y φ χ νS carries dark charge 2Q, so the φ→χ+νS decay and the entire ΔNeff mechanism are not gauge-invariantly defined.","rationale":"The reader's weakest-assumption analysis correctly identifies the gauge non-invariance of the central Yukawa interaction as the most load-bearing defect. If Eq. (1) is read literally, with equal U(1)_D charges for χ and φ and a neutral νS, the decay ϕ→χ+νS is simply not gauge invariant, and the paper's hybrid relic mechanism and the associated ΔNeff prediction collapse. This is not a matter of consensus or numerical uncertainty; it is an internal consistency failure at the defining vertex. I agree with the reader that the fix is straightforward—writing the conjugate contraction or reversing one charge assignment—so the appropriate disposition remains conditional rather than rejection. I also note the secondary free-streaming problem in Table I and Eq. (28), which independently weakens the structure-formation claim but does not by itself destroy the central mechanism; it reinforces the need for a revised manuscript. The reader's proposed remedy of 'opposite charges' is one valid repair only if the operator remains as written without a bar; if the intended operator is y ϕ \\barχ νS, equal charges are already correct. Because the printed text lacks the bar, the concern lands as stated, and I mark agreement with the reader's identification of the weakest assumption. The concrete test of writing out the gauge transformation and checking invariance, followed by a correction of the operator and a recalculation of the affected benchmarks, would settle whether this is a typographical omission or a substantive flaw. The paper's numerical scans are not public, so the recommended conditional acceptance should also carry a request for release of the scan code or benchmark data.","tokens_in":19204,"tokens_out":30247,"duration_ms":289692,"concrete_test":"Check gauge invariance of Eq. (1) explicitly: assign U(1)_D charges Qχ=Qφ=Q, QνS=0 and apply the gauge transformation to y ϕ χ νS; the term changes by e^{2iQα} and is forbidden. Then repeat the check for y ϕ \\barχ νS; if that is the intended operator, equal charges are consistent and the manuscript should state the contraction explicitly. Additionally, recompute Table I BP1 using the full free-streaming integral in Eq. (26) with v(a)=1/sqrt(1+(a/anr)^2); if λfs remains above 0.1 Mpc, that benchmark must be removed or re-parameterized before the claimed consistency with structure formation is accepted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II states that χ and φ carry equal U(1)_D charge while νS is neutral, and Eq. (1) then contains the term (y ϕ χ νS + h.c.). Under a U(1)_D transformation with Qχ=Qφ=Q and QνS=0, the field product ϕ χ νS transforms with charge 2Q, so the operator is forbidden by the unbroken symmetry that the paper invokes to stabilize χ. If the intended operator is actually y ϕ \\barχ νS, the missing bar would make the equal-charge assignment gauge-invariant, but as printed the interaction is inconsistent with the stated charge assignments. This vertex is the sole source of late-time non-thermal dark matter production and of the dark radiation νS; without it, ϕ does not decay to χ+νS, the thermal underabundance is not compensated, and ΔNeff receives no contribution from this mechanism. The central claim therefore depends on an interaction that the paper's own charge assignments exclude. The repair is minimal—restore the conjugate contraction or flip one charge—but the submitted text is internally inconsistent at the defining vertex. A related but secondary internal inconsistency is that benchmark BP1 in Table I has λfs=0.256 Mpc, violating the paper's own λfs<0.1 structure-formation bound, and the approximate free-streaming integral in Eq. (28) appears to underestimate λfs relative to the full expression in Eq. (26).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a dark U(1)_D extension of the SM containing a GeV-scale Dirac fermion DM χ, a light Stueckelberg vector X that mediates χ self-interactions, a complex singlet scalar φ, and a sterile neutrino νS. With U(1)_D left unbroken, χ is stable, and the relic density is generated in two stages: thermal freeze-out of χχ→XX, which is underabundant in the SIDM-relevant parameter region, plus non-thermal production through the late decay φ→χ νS occurring after BBN and before the CMB epoch. The same decays produce the dark radiation νS, giving ΔNeff in the range 0.01–0.2, claimed to be compatible with Planck and within reach of SPT-3G/CMB-S4/CMB-HD, while a scan over gd, mφ, mχ, and y imposes the observed relic density, the SIDM self-interaction cross-section, and the free-streaming bound λfs < 0.1 Mpc. Three benchmark points are given in Table I, and the paper includes detailed appendices for the cross-sections, Boltzmann equations, and Stueckelberg mass generation.","tokens_in":19649,"tokens_out":50820,"duration_ms":445795,"significance":"The setup is timely and the hybrid production mechanism is standard and clearly articulated; the paper is honest about the tension between efficient annihilation into light mediators and the resulting underabundant thermal relic, and it documents the model in unusual detail (cross-sections, Boltzmann equations, decay widths, direct detection analysis). Its concrete falsifiable predictions — ΔNeff within reach of next-generation CMB experiments and spin-independent scattering targets in Section V — are a genuine strength. However, the quantitative demonstration of the central claim is not currently reliable: the defining Yukawa interaction is written in a way that violates the stated unbroken U(1)_D; the free-streaming evaluation in Section VI appears to be wrong by orders of magnitude, so the claimed compatibility of the benchmarks (and the scan of Fig. 15) with λfs < 0.1 Mpc is not established; and Table I is internally inconsistent. These problems are fixable in principle but require a substantive recomputation, so the paper is not yet publishable in its present form.","major_comments":[{"comment":"As printed, the interaction (y φ χ νS + h.c.) is not gauge invariant under the charge assignments stated in Section II, namely Qχ = Qφ and QνS = 0: the field product carries net dark charge 2Q and is forbidden by the unbroken U(1)_D that the paper invokes to stabilize χ. This vertex is the sole source of the late-time non-thermal DM production and of the dark radiation νS, so the central decay mechanism is not consistently defined as written. The likely intended operator is y φ \\barχ νS, which is invariant under equal charges; please correct the printed operator (or flip one of the charge assignments) and state the charges explicitly.","section":"Section II, Eq. (1)"},{"comment":"The approximate free-streaming evaluation in Eq. (28) is internally inconsistent. The final expression vkick (Γφ)^{-1} a_d ln(a_eq/a_d) does not follow from the displayed integral: inserting H(a) = H0√Ωr a^{-2} and Γφ = H(a_d) = H0√Ωr a_d^{-2} gives λfs = vkick a_d ln(a_eq/a_d)/(H0√Ωr), not the printed result, which contains an extra factor a_d^2. Moreover, the integral drops the relativistic stage a_d < a < a_nr, where v ≈ 1, and the matter-era tail, both of which are required by Eq. (26). Evaluating Eq. (26) with the kinematics of BP2 (a_d = 6.84e-6, a_nr = 1.56e-4, vkick = 0.999) gives λfs ~ O(100) Mpc: the relativistic segment alone contributes roughly a_nr/(H0√Ωr) ≈ 70 Mpc, and the non-relativistic radiation-era segment contributes another ~45 Mpc, versus the quoted 0.006 Mpc. The paper's central claim that the benchmarks and the scan of Fig. 15 respect λfs < 0.1 Mpc therefore needs to be re-established with the full integral.","section":"Section VI, Eqs. (26)–(28), Table I"},{"comment":"BP1 is inconsistent with the paper's own structure-formation constraint. The row reports λfs = 0.256 Mpc, which violates the bound λfs < 0.1 Mpc quoted in Section VI (following Refs. [66–68]), and it also has a_d = 5.2e-4 > a_eq = 2.94e-4 and a_nr > a_eq, contradicting the ordering a_d << a_nr < a_eq used to justify Eqs. (26)–(28). The statement in Section VI that the three benchmark scenarios of Table I illustrate that 'it is indeed possible to satisfy the structure formation constraints' is therefore not correct as printed; BP1 should be removed or replaced with a point that genuinely satisfies all stated constraints.","section":"Table I, BP1 row"},{"comment":"The quoted Yukawa couplings are inconsistent with the quoted decay epochs. Section VI states that 'the decay happens when Γφ = H(a_d)'; using the width Γφ = y^2 mφ (1 - mχ^2/mφ^2)^2/(16π), BP2 (y = 8.1e-16, mφ = 46.56 GeV) gives Γφ ≈ 9e-7 s^{-1}, which equals H at T ≈ 1.2 keV (a ≈ 2e-7), whereas Table I lists a_d = 6.84e-6, where H(a_d) ≈ 5-8e-10 s^{-1}, i.e., Γφ/H ≈ 10^3. The same discrepancy (Γφ/H(a_d) of order 10^3) occurs for BP1 and BP3. If a_d is intended to be an effective epoch of maximal decay or energy injection rather than the Γφ = H epoch, its definition should be stated; as it stands, the benchmark table and the associated ΔNeff and λfs values cannot be reproduced from the formulas given in the paper.","section":"Table I vs. Eq. (16) and Section VI"},{"comment":"The φφ → XX annihilation cross-section is quoted inconsistently: Section III.B states ⟨σv⟩_{φφ→XX} ≃ 6πα_d/mφ^2, while the derivation in Appendix D, Eq. (D1), gives ⟨σv⟩ = 6 g_d^4/(16π mφ^2) = 6πα_d^2/mφ^2. The two expressions differ by a factor α_d ≈ 10^{-4}–10^{-3} for the couplings considered. Since the thermal abundance Yφ — and hence the non-thermal DM component and ΔNeff, both of which are produced from φ decays — is controlled by this cross-section, the correct expression must be identified and the scans (Figs. 5, 8–10, 15) recomputed consistently.","section":"Section III.B and Appendix D, Eq. (D1)"}],"minor_comments":[{"comment":"Several displayed formulas have lost their Dirac conjugation bars in the typeset version, e.g., the mass term 'mχχχ' and the operator 'yφχνS' in Eq. (1), and the phrase 'χχ and χχ interactions are repulsive' in Section II.A should presumably read χ\\barχ and \\barχ\\barχ; please ensure the published version typesets correctly, as these details matter for the gauge-invariance discussion in the first major comment.","section":"Sections II and II.A"},{"comment":"The text states that for most calculations a 'simplified form of the Boltzmann equation that does not require tracking the co-moving density of X' is adopted, but no numerical justification is given. Since the X boson decays to the SM only through a very small kinetic mixing (ε down to 10^{-10}), the lifetime of X can be comparable to BBN timescales and a residual X abundance could, in principle, alter the χ relic density or inject energy; please state the explicit check that validates dropping YX.","section":"Appendix F"},{"comment":"The definition of ΔNeff in Eq. (17) should specify the epoch at which ρνS/ρνL is evaluated and the temperature convention used (photon vs. neutrino temperature); as written, the notation 'ρνS/ρνL|_{TCMB}' is ambiguous about the integration of the Boltzmann equation (18) and about how the SM value N_eff^SM = 3.045 enters.","section":"Section IV, Eq. (17)"},{"comment":"Section IV states that MX is restricted to 1 MeV or below 'ensuring that it can decay only into SM neutrinos', but Fig. 2 displays results for MX = 10 MeV, for which X → e+e- is kinematically open; please clarify that all benchmark and ΔNeff results use MX ≤ 1 MeV, or comment on the e+e- constraints for the 10 MeV case.","section":"Section IV and Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The referee report is self-contained; in confidence I add two observations. First, the apparent order-of-magnitude error in the free-streaming evaluation (Section VI) is the main obstacle to publication: if it survives in the revision, the claimed simultaneous compatibility of observable ΔNeff with λfs < 0.1 Mpc is not demonstrated and the central phenomenological result would be unsupported. Second, the reference list leans heavily on the authors' own previous work; I am not treating this as a flaw in itself, but an independent cross-check of the numerical pipeline (e.g., reproducing one benchmark from the published formulas, or releasing the code) would substantially increase confidence in the revised numbers."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThe punchline: this is an incremental but potentially useful SIDM model connecting a GeV-scale dark matter candidate to an observable ΔNeff, but the submitted text has a load-bearing internal inconsistency at the defining vertex. Section II assigns χ and φ equal U(1)_D charge with νS neutral, which makes the Yukawa term y φ χ νS in Eq. (1) carry net dark charge 2Q, violating the unbroken gauge symmetry. That vertex is the only source of the late decay that replenishes DM and produces dark radiation; without it the mechanism doesn't run. The repair is minimal (missing conjugate on χ or flip a charge), but as printed the model isn't gauge-invariant.\n\nWhat's actually new: the hybrid freeze-out plus late scalar decay is present in the authors' own prior work [27–30]; the new bits are the Stueckelberg mass for the mediator and the specific scan against ΔNeff. That's a modest extension, but the paper does it carefully: the Boltzmann equations are laid out, direct detection and structure formation constraints are included, and the free-streaming length is discussed. If the gauge issue is fixed, the ΔNeff predictions are plausible.\n\nSoft spots beyond the charge assignment: benchmark BP1 in Table I has ad > aeq (decay during matter domination) and λfs = 0.256 Mpc, violating the paper's own λfs < 0.1 Mpc bound. The approximate free-streaming integral in Eq. (28) assumes ad << aeq and may underestimate λfs in that regime. Also, the numerical scans are not public, so only representative points can be checked; with free parameters gd, mχ, mφ, y, MX, ΔNeff is a consistency window rather than a sharp prediction.\n\nWho this is for: model-builders working on SIDM and CMB spectral distortions. It deserves a serious referee—the mechanism is standard enough and the topic is active—but the authors need to correct the charge assignment and revisit BP1 before publication. I wouldn't cite it as submitted, but I'd encourage the editor to send it to review with those requests.\n\nBest,\n[You]","headline":"A fixable but load-bearing gauge-invariance error in Eq. (1) and a benchmark violating its own bound undercut an otherwise plausible SIDM increment to ΔNeff.","tokens_in":20159,"tokens_out":4832,"would_cite":false,"duration_ms":41243,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A GeV-scale self-interacting dark matter candidate can match the observed relic abundance while producing an observable ΔNeff through late scalar decay.","keywords":["self-interacting dark matter","dark radiation","observable ΔNeff","gauged U(1)_D extension","Stueckelberg mechanism","non-thermal dark matter production","thermal relic underabundance","small-scale structure anomalies"],"falsifier":"Check the dark charge of the decay term $y\\,\\varphi\\,\\chi\\,\\nu_S$ under the model's own charge table: the paper assigns the same charge to $\\chi$ and $\\varphi$ and zero to $\\nu_S$, so the operator carries net charge $2q_D$ and is forbidden by the unbroken symmetry; if the charge is not cancelled, both the non-thermal dark matter and the dark radiation are absent and the quoted $\\Delta N_{\\rm eff}$ vanishes, while any repaired assignment changes the freeze-out and decay dynamics enough to require recomputation.","tokens_in":19035,"feed_emoji":"🌌","tokens_out":23193,"duration_ms":193586,"temperature":0.7,"pith_summary":"This paper tries to show that the same dark sector can solve both the small-scale puzzles of cold dark matter and the underabundance problem of light thermal relics, while leaving a measurable trace in the cosmic radiation content. The model adds a GeV-scale fermion $\\chi$ that scatters off itself through a light dark vector boson $X$, and a heavier scalar $\\varphi$ that decays after big bang nucleosynthesis but before the cosmic microwave background epoch into $\\chi$ and a dark radiation fermion $\\nu_S$. This late decay tops up the thermally underabundant dark matter and injects the dark radiation that shifts the effective number of relativistic species by $\\Delta N_{\\rm eff}$ in the range roughly 0.01 to 0.2. A sympathetic reader would take the paper as establishing a proof of concept: strong self-interactions, correct relic abundance, and an observable cosmological signature can coexist in one minimal model.","feed_headline":"One scalar decay tops up dark matter's relic abundance","feed_subtitle":"The same decay emits dark radiation, giving an observable ΔNeff within next-generation CMB reach.","key_machinery":"The central mechanism is the hybrid relic: thermal freeze-out sets the underabundance and a late scalar decay fills the gap. The scalar $\\varphi$ stays in equilibrium with the Standard Model bath through its Higgs portal coupling $\\lambda_{h\\varphi} \\gtrsim 10^{-5}$, then decouples and decays with width $\\Gamma_\\varphi = y^2 m_\\varphi/(16\\pi)\\,(1 - m_\\chi^2/m_\\varphi^2)^2$. The same coupling $y$ controls both the injection of $\\chi$ and the energy density of $\\nu_S$, which is evolved through a Boltzmann equation; because $\\rho_\\varphi$ grows relative to radiation after $\\varphi$ becomes non-relativistic, a smaller $y$ means a later decay and a larger $\\Delta N_{\\rm eff}$. The light vector $X$, with its Stueckelberg mass and kinetic mixing $\\epsilon$, supplies an attractive exponential potential whose transfer cross-section $\\sigma_T$ defines the self-interaction regime and also mediates the direct-detection rate.","core_discovery":"On its own terms, the paper's central claim is that a self-interacting dark matter candidate at the GeV scale can avoid the standard thermal-relic obstruction. The dark sector consists of a fermion $\\chi$, a light vector mediator $X$ whose mass is generated through the Stueckelberg mechanism (a mass for a gauge boson without a scalar vacuum expectation value), a singlet scalar $\\varphi$, and a dark radiation fermion $\\nu_S$; $\\chi$ freezes out with a relic density below the observed value because annihilations into the light mediator are efficient. The scalar $\\varphi$ freezes out with a larger abundance and later decays through $\\varphi \\to \\chi + \\nu_S$, supplying the missing dark matter and simultaneously creating the dark radiation that produces $\\Delta N_{\\rm eff}$. The paper computes the coupled Boltzmann evolution and finds parameter points, with $\\Delta N_{\\rm eff}$ from about 0.01 to 0.2, that agree with current CMB measurements and lie within the projected reach of future CMB experiments while respecting direct detection and free-streaming limits. The intended message is that small-scale structure anomalies, the light-thermal-relic underabundance, and the search for extra radiation are three facets of one mechanism.","pith_inferences":["The charge-assignment issue is repairable, but any repair (giving $\\nu_S$ a dark charge or flipping the charge of $\\varphi$) changes the dark-sector dynamics, so the benchmark numbers would shift even if the qualitative story survives.","The thermal-underabundance-plus-late-injection pattern is generic: any model with a strongly annihilating light dark matter candidate and a long-lived heavier scalar decaying into dark matter plus a light fermion will produce a similar $\\Delta N_{\\rm eff}$ signature, making future CMB measurements a broad probe of non-thermal production.","The authors' alternative completion in which the scalar decays to dark matter and active neutrinos would blur the cosmological signature, because active neutrinos are already counted in the baseline $N_{\\rm eff}$; distinguishing that branch from the $\\nu_S$ branch would require more than a total $\\Delta N_{\\rm eff}$ measurement."],"forward_implications":["The model removes the main obstruction for GeV-scale self-interacting dark matter: the thermal underabundance from efficient annihilation into light mediators is compensated by the late scalar decay, so the observed relic density and a self-interaction cross section near 1 cm²/g can coexist.","The same decay injects a dark radiation component, giving $\\Delta N_{\\rm eff}$ from roughly 0.01 to 0.2 in the scanned region of parameter space, which current CMB bounds allow and next-generation CMB experiments can test.","Direct detection already closes the branch with kinetic mixing $\\epsilon = 10^{-9}$ for dark matter masses above about 3 GeV; reducing the mixing to $10^{-10}$ keeps masses up to about 6 GeV viable, so near-future detectors will cover the rest.","The free-streaming constraint $\\lambda_{\\rm fs} < 0.1$ Mpc selects a definite band of decay parameters, so relic abundance, $\\Delta N_{\\rm eff}$, and structure formation jointly pin down the scalar mass, dark matter mass, and coupling $y$."],"supporting_citations":[{"why":"Sets the self-interaction cross-section target around 1 cm²/g that the model must produce.","marker":"[4–7]"},{"why":"Supplies the transfer cross-section for the exponential potential used to compute the dark matter self-interaction rate.","marker":"[11]"},{"why":"Establishes the few-GeV lower bound on thermal WIMP relics, the problem the non-thermal top-up is designed to solve.","marker":"[18,19]"},{"why":"Provides the measured bound on the effective number of relativistic species that the predicted excess must satisfy.","marker":"[32]"},{"why":"Gives the Stueckelberg mass-mixing prescription that gives the light vector mediator its mass.","marker":"[36]"},{"why":"Fixes the Standard Model baseline effective number of relativistic species near 3.045 against which the excess is defined.","marker":"[38–40]"},{"why":"Provides the thermal-averaged energy transfer rate used in the Boltzmann equation for the dark radiation density.","marker":"[44]"},{"why":"Impose the free-streaming length bound that restricts the allowed decay parameters.","marker":"[66–68]"},{"why":"Gives the free-streaming length formula used to check the structure formation constraint.","marker":"[69–71]"}],"fun_headline_variants":["Scalar decay tops up DM, emits detectable ΔNeff","Late scalar decay: DM top-up and CMB radiation","Dark scalar's late decay fills DM and lights CMB","One decay: DM from underabundance, ΔNeff in CMB"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole mechanism depends on the scalar being able to decay into the dark matter particle plus the dark radiation particle, but under the charge assignments stated in the paper that decay is forbidden by the unbroken symmetry of the dark force; unless the assignment is corrected, the late production that generates the relic top-up and the extra radiation does not exist.","fun_headline_variants_meta":{"raw":{"variants":["Scalar decay tops up DM, emits detectable ΔNeff","Late scalar decay: DM top-up and CMB radiation","Dark scalar's late decay fills DM and lights CMB","One decay: DM from underabundance, ΔNeff in CMB"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000603,"raw_usage":{"total_tokens":2870,"prompt_tokens":1057,"completion_tokens":1813,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":673,"completion_tokens_details":{"reasoning_tokens":1741}},"tokens_in":673,"tokens_out":1813,"duration_ms":12682,"temperature":1.0,"reasoning_tokens":1741,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:53:53.941290+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the dark charge of the decay term $y\\,\\varphi\\,\\chi\\,\\nu_S$ under the model's own charge table: the paper assigns the same charge to $\\chi$ and $\\varphi$ and zero to $\\nu_S$, so the operator carries net charge $2q_D$ and is forbidden by the unbroken symmetry; if the charge is not cancelled, both the non-thermal dark matter and the dark radiation are absent and the quoted $\\Delta N_{\\rm eff}$ vanishes, while any repaired assignment changes the freeze-out and decay dynamics enough to require recomputation.","supporting_citations":[],"review_version":1}