{"id":"6d6ba8d5-68b5-4397-aa32-cb2b720b80f0","arxiv_id":"2512.18186","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A JUNO-like detector could constrain the neutrino–ultralight-dark-matter oscillation-modulation parameters to ηΔ21 < 2.5×10^-2 and ηΔ31 < 5×10^-3 at 90% CL.","lead":"This paper estimates how well a JUNO-like neutrino detector could detect ultralight dark matter that changes neutrino masses over time. It finds the detector could set new limits on the dark-matter–neutrino coupling, though the effect on mass-ordering sensitivity is modest.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (6) time-averaging truncation is unquantified and can be ~20% of the signal at the ηΔ31 limit, so the headline limits and CMB comparison are not yet quantitatively secure.","rationale":"The reader identified the detector/systematics modeling as the weakest assumption and mentioned the time-averaging truncation only in passing. My read agrees that the detector model is a conditional assumption, but the more load-bearing point for the central quantitative claim is the unquantified truncation in Eq. (6), because it directly enters every sensitivity number before any detector/systematics refinement is applied. The claimed limits and the 'surpassing CMB' statement depend on the η^2 signal shape; at the relevant x_E values the omitted O(x_E^4 η^4) term is not negligible. This is a concrete, checkable issue, not a disagreement with consensus: the exact Bessel-function average is the standard result, and the question is whether the paper's approximation is accurate enough for the stated precision. The paper is otherwise a clean sensitivity study, and the concern is addressable without undermining the overall approach. I therefore keep the reader's CONDITIONAL verdict unchanged: the central claim is plausible but not yet firmly established at the claimed precision.","tokens_in":12437,"tokens_out":20831,"duration_ms":193460,"concrete_test":"Rerun the GLoBES simulation replacing the truncated Eq. (6) with the exact time average 1/2 − 1/2 cos(2x_E) J0(4x_Eη) for each sin^2 term in the reactor ν̄e survival probability, keeping all detector/systematic assumptions unchanged. If the resulting 90% CL limits on ηΔ21 and ηΔ31 shift by less than ~5%, the concern is resolved; if they shift by more than ~10–20%, the headline limits and the Yukawa-coupling comparison with CMB (Fig. 3) must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central signal model, Eq. (6), expands the time-averaged oscillation probability as sin^2(x_E) + 2 x_E^2 η^2 cos(2x_E) + O(x_E^4 η^4), keeping only the η^2 term. For the quoted ηΔ31 90% limit (0.5×10^-2), low-energy bins near the 1.8 MeV threshold have x_E = Δm^2 L/(4E) ≈ 90 rad, so the expansion parameter 2x_E η ≈ 0.9. The first omitted term is ≈ 2 x_E^4 η^4, which is about 0.09 of the probability amplitude, i.e., ~20% of the retained η^2 signal in those bins. The same issue affects ηΔ21 less severely but is still unquantified. Since the Δχ^2 is quadratic in the signal, a 10–20% shape error can shift the 90% CL limits by a comparable amount. The manuscript neither bounds this truncation error nor compares with the exact Bessel average (1/2 − 1/2 cos(2x_E) J0(4x_Eη)). Thus the quoted ηΔjk limits—and the factor-of-3 'surpassing CMB' statement in Fig. 3—are not yet quantitatively supported at the stated precision, independent of the detector/LSNL caveats in Section V.D.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies ULDM coupled to neutrinos through a Yukawa interaction that modulates neutrino mass-squared differences. In the regime where the ULDM oscillation period is much shorter than the detector exposure, the authors use a time-averaged expansion (Eq. 6) to obtain spectral distortions parameterized by ηΔ21 and ηΔ31. Using a GLoBES simulation of a JUNO-like liquid scintillator detector (20 kton, 52.5 km, 6.5 years × 26.6 GWth, 90% efficiency), they derive projected 90% CL sensitivities ηΔ21 ≲ 2.5×10^-2 and ηΔ31 ≲ 0.5×10^-2, translate these into bounds on neutrino-ULDM Yukawa couplings, and study correlations with Δm^2, effects on oscillation parameter determination, and mass-ordering sensitivity. The paper concludes that a large liquid scintillator detector can set limits surpassing current CMB-derived bounds and complement DUNE and ESSnuSB projections.","tokens_in":12842,"tokens_out":16097,"duration_ms":145967,"significance":"The paper is a useful and timely projection study for a major upcoming experiment. It uses standard tools (GLoBES, Huber-Mueller fluxes, pull-based χ^2) and clearly documents the assumed detector parameters, systematics, and priors. The qualitative conclusions — that ULDM effects can mimic energy smearing, that ηΔ31 has a mild impact on mass-ordering sensitivity, and that a JUNO-like detector is competitive in this channel — are of interest to the neutrino and dark matter communities. However, the quantitative headline limits are not yet secure because of an unquantified truncation error in the signal model and an inconsistency in the density-dependent scaling of the Yukawa bounds. If these are corrected, the paper would be a solid contribution.","major_comments":[{"comment":"The signal model in Eq. (6) truncates the time-averaged probability at O(x^4 η^4). For the quoted ηΔ31 90% limit (0.5×10^-2), the lowest-energy bins near 1.8 MeV have x ≈ 94, so 2xη ≈ 0.94 and the first omitted term, −2x^4η^4 cos2x, has magnitude ≈0.1, i.e. roughly 20–25% of the retained 2x^2η^2 term in those bins. Because Δχ^2 is quadratic in the signal, a shape error of this size can shift the projected limit by a comparable amount. The exact time average is 1/2 − 1/2 cos(2x) J0(4xη); the paper should either use this expression or quantify the truncation error and demonstrate that the 90% limits are robust. This is load-bearing for the ηΔjk limits and for the comparison to CMB constraints in Fig. 3.","section":"Eq. (6), Section V.A"},{"comment":"The density bookkeeping is inconsistent. Eq. (5) depends on the local ULDM density, but Section V.A defines ρφ = 0.1ρDM ≃ 10^-12 eV^4 with ρDM,⊙ ≃ 10^5 ρDM, so the local ULDM density would be ρφ,⊙ ≈ 0.1ρDM,⊙ ≈ 10^-7 eV^4, not 10^-12 eV^4. Eq. (12) contains a factor ρφ/(0.1ρDM,⊙), which evaluates to 10^-5 under these definitions, yet the numerical prefactors 3×10^-22 and 4×10^-22 are correct only if ρφ,⊙ = 0.1ρDM,⊙ (factor 1). Moreover, from Eq. (5), for fixed η the scaling is y ∝ 1/√ρφ,⊙, not y ∝ ρφ as printed. This discrepancy propagates directly into Fig. 3 and into the headline statement that the projected JUNO bounds 'surpass' CMB constraints. The correct local density and scaling should be clarified and used consistently.","section":"Eq. (12), Section V.A, Fig. 3"},{"comment":"Section V.D states that a precise quantitative assessment requires a detailed implementation of liquid-scintillator non-linear effects, yet the abstract and conclusions present the 90% CL limits and the comparison with CMB constraints without this caveat. The same simplified detector model underlies the central limits in Section V.A. The authors should either implement the non-linear response or explicitly state in the abstract and conclusions that the reported limits are preliminary projections based on an idealized detector model.","section":"Section V.D and Conclusions"}],"minor_comments":[{"comment":"Typo: 'liquid scintillator non-nonlinearities' should be 'non-linearities'.","section":"Section VI"},{"comment":"The notation '[eV^4]' inside the denominator ratio is unusual and can be misread as a dimensional factor. Use dimensionless ratios throughout, e.g. ρφ/(0.1ρDM,⊙) with the appropriate normalization.","section":"Eq. (12)"},{"comment":"The sentence 'for a scalar field modulation period τφ ≈ 1 year' does not reproduce the lower bound 3.0×10^-23 eV in Eq. (11); with τφ = 1 year one obtains mφ ≈ 1.3×10^-22 eV. Please clarify whether Eq. (11) results from the requirement τφ ≪ τexp or from an explicit τφ ≈ 1 year example.","section":"Section V.A, Eq. (11)"},{"comment":"The legends in Fig. 5 repeat 'SM+ηΔ21=2x10^-2' several times; the figure would be easier to read with a single legend entry per curve.","section":"Fig. 5"}],"recommendation":"major_revision","confidential_remarks":"The paper is a reasonable projection study, but the two central quantitative claims — the ηΔjk limits and the Yukawa coupling bounds — depend on technical issues that must be corrected before publication. The truncation issue in Eq. (6) is readily fixable by using the exact Bessel-averaged expression. The density scaling inconsistency in Eq. (12) appears to be a bookkeeping error rather than a deep flaw, but it affects Fig. 3 and the paper's main 'surpassing CMB' statement. I recommend major revision and would want to see the revised version before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nYou can read this as a standard, careful sensitivity study: a JUNO-like detector, ULDM-induced time-averaged oscillation distortions, and projected 90% limits on ηΔ21 and ηΔ31. The formalism is not new — it comes from Krnjaic/Machado/Necib and Losada et al., plus the authors’ own earlier work — but the paper adds a full GLoBES simulation, updated detector assumptions, and a look at correlations with oscillation parameters and mass-ordering sensitivity. That is a legitimate extension, and the detector simulation is described in enough detail that the numbers are checkable in principle. Credit is due for being explicit about the caveat on liquid-scintillator non-linearities and for framing the results as projections, not claims on data.\n\nThe main soft spot is Eq. (6). The time-averaged probability is expanded to second order in η, and the truncation error is never quantified. The stress-test math holds up: at the ηΔ31 limit, low-energy bins have x ≈ 90 and 2xη ≈ 0.9, so the first omitted term is about 20% of the retained η² signal. The same issue barely affects ηΔ21. Since the omitted term is a smooth function of energy, it can bias the spectral shape and shift the 90% limit by maybe 10% — not enough to overturn the result, but enough that the stated limits and the “surpassing CMB” line in Fig. 3 are overstated as written. The fix is trivial: use the exact Bessel-function average, or at least bound the next term. A referee should ask for that.\n\nA few smaller things. The comparison with the earlier JUNO projection in Ref. [10] is not quantitative — the paper says a preliminary assessment was done there but never says how much better this simulation is. No code or data are released, so the simulation is not independently reproducible without reimplementing it. The mass-ordering degradation (Δχ² ≈ 2.5) is mild, and the authors themselves label it qualitative.\n\nOverall: the central claim — that JUNO can reach ηΔ31 ≲ 5×10⁻³ and ηΔ21 ≲ 2.5×10⁻² — is plausible and worth taking seriously. The paper is not a breakthrough, but it is a solid projection that deserves a serious referee. I would send it to review, with the request that the averaging expansion be checked and the CMB comparison be rephrased.\n\nBest,\n[Your name]","headline":"A competent, incremental JUNO-like sensitivity projection for neutrino-ULDM couplings; the headline limits are plausible, but the unquantified truncation of the averaging expansion is a real soft spot and the CMB-surpassing claim is somewhat overstated.","tokens_in":13319,"tokens_out":3820,"would_cite":true,"duration_ms":36986,"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":"A 20-kiloton liquid-scintillator reactor detector can set neutrino–ultralight-dark-matter coupling limits that beat current CMB bounds.","keywords":["ultralight scalar dark matter","neutrino oscillations","reactor antineutrinos","liquid scintillator detector","Yukawa coupling bounds","mass-squared difference modulation","neutrino mass ordering","time-averaged oscillation"],"falsifier":"Fit the real detector's observed energy spectrum with the same χ² and priors once the data are available: if the data-driven 90% CL interval for ηΔ31 is not bounded above by about 0.5×10^-2 (or for ηΔ21 by about 2.5×10^-2), the paper's projected sensitivity is contradicted. An independent calculation that replaces the fixed Gaussian 3% resolution with a full non-linear liquid-scintillator response model, and that yields limits weaker by more than the quoted uncertainties, would also falsify the quantitative claim.","tokens_in":12354,"feed_emoji":"⚛️","tokens_out":6788,"duration_ms":70081,"temperature":0.7,"pith_summary":"The paper argues that a large liquid-scintillator detector watching reactor antineutrinos—of the type planned for the JUNO experiment—can serve as a dark-matter probe. If ultralight scalar dark matter couples to neutrinos, it makes the neutrino mass-squared differences oscillate, and when that oscillation is faster than the detector's integration time, the effect survives as a time-averaged smearing of the neutrino oscillation probability. That smearing looks like an extra energy-resolution term, so it can be constrained from the shape of the reactor-neutrino spectrum. Under a simulated 20-kton, 52.5-km detector configuration, the projected 90% CL sensitivities are ηΔ21 ≲ 2.5×10^-2 and ηΔ31 ≲ 0.5×10^-2, which translate into Yukawa couplings around 3×10^-22 and 4×10^-22 at a scalar mass of 10^-21 eV, surpassing CMB-derived limits and complementing DUNE and ESSnuSB projections. The paper also shows that such couplings, if present near these values, would mildly degrade the precision of Δm²31 and the sensitivity to the neutrino mass ordering.","feed_headline":"A reactor-neutrino detector can beat CMB limits on ultralight dark matter","feed_subtitle":"Projected 90% limits reach Yukawa couplings near 3e-22, a window CMB bounds cannot cover.","key_machinery":"The averaging identity ⟨sin²[x(1+2η sin(mφt))]⟩ = sin²x + 2x²η² cos(2x) + O(x⁴η⁴), obtained by integrating over one ULDM period τφ. This operation converts a rapidly oscillating modification of the mass-squared differences into a static, energy-dependent smearing of the oscillation probability—and it is why the effect survives time averaging and can be searched for as a distortion in the measured antineutrino spectrum. The analysis also relies on the hierarchy τν ≪ τφ ≪ τexp, which delimits the ULDM mass window of roughly 10^-23 to 10^-11 eV for the 52.5-km baseline and 6.5-year exposure.","core_discovery":"The central quantitative claim is that a 20-kton liquid-scintillator detector at 52.5 km from a reactor, with 6.5 years of 26.6 GWth exposure, 90% efficiency, 3% energy resolution, and standard background and systematic assumptions, can set 90% CL upper limits of ηΔ21 ≲ 2.5×10^-2 and ηΔ31 ≲ 0.5×10^-2 on the ULDM modulation parameters. These parameters measure the fractional time-averaged smearing of the solar and atmospheric mass-squared splittings induced by a coherently oscillating ultralight scalar field. The paper further shows that the same couplings, if present at the level of the projected sensitivity, would appear as spectral distortion that mimics energy smearing, producing only mil","pith_inferences":["The paper leaves implicit that better energy resolution than 3% would sharpen the ηΔ31 reach more than the ηΔ21 reach, because the atmospheric term's oscillation phase x = Δm² L/(4E) grows with the larger splitting; a dedicated experiment could exploit that asymmetry.","The quoted Yukawa numbers assume ULDM is 10% of dark matter; if ULDM is the full dark-matter density, the same η limits translate to couplings roughly three times smaller, and if it is a smaller subcomponent, the bounds weaken accordingly.","The same time-averaged modulation formalism could be applied to other long-baseline neutrino sources or to atmospheric neutrinos at this detector, a direction the paper mentions but does not develop numerically."],"forward_implications":["A detector of this size and baseline can set 90% CL bounds on neutrino–ULDM couplings that beat current CMB limits in the mass window roughly 10^-23 to 10^-11 eV, without needing new apparatus.","The ULDM signal is a spectral smearing; if present at the projected level, it will look like degraded energy resolution and could bias a standard three-neutrino fit of Δm²31 unless included in the fit.","The neutrino mass-ordering sensitivity in this type of detector would be reduced by Δχ² ≈ 1–2.5 for η values near the projected 1σ reach.","The projected Yukawa limits complement DUNE and ESSnuSB, which probe different baseline and energy regimes."],"fun_headline_variants":["Liquid scintillator detector probes ultralight dark matter","Neutrino oscillations reveal ultralight dark matter signals","20-kton detector sets limits on dark matter-induced neutrino smearing","Reactor neutrinos hunt ultralight scalar dark matter","Large scintillator detector beats CMB on dark matter sensitivity"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The quantitative limits hinge on the simulated detector model—specifically, that a 20-kton, 52.5-km liquid-scintillator detector with 3% energy resolution, the assumed reactor flux, backgrounds, and systematic uncertainties, faithfully represents the real experiment; the paper itself notes in Section V.D that a precise assessment requires implementing liquid-scintillator non-linear effects, so these numbers are not final.","fun_headline_variants_meta":{"raw":{"variants":["Liquid scintillator detector probes ultralight dark matter","Neutrino oscillations reveal ultralight dark matter signals","20-kton detector sets limits on dark matter-induced neutrino smearing","Reactor neutrinos hunt ultralight scalar dark matter","Large scintillator detector beats CMB on dark matter sensitivity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000733,"raw_usage":{"total_tokens":3106,"prompt_tokens":722,"completion_tokens":2384,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":466,"completion_tokens_details":{"reasoning_tokens":2311}},"tokens_in":466,"tokens_out":2384,"duration_ms":18706,"temperature":1.0,"reasoning_tokens":2311,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T15:03:59.383905+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fit the real detector's observed energy spectrum with the same χ² and priors once the data are available: if the data-driven 90% CL interval for ηΔ31 is not bounded above by about 0.5×10^-2 (or for ηΔ21 by about 2.5×10^-2), the paper's projected sensitivity is contradicted. An independent calculation that replaces the fixed Gaussian 3% resolution with a full non-linear liquid-scintillator response model, and that yields limits weaker by more than the quoted uncertainties, would also falsify the quantitative claim.","supporting_citations":[],"review_version":1}