{"id":"c750a1bc-29d9-4035-b78e-d4219ab7aa0e","arxiv_id":"2411.19836","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Xenon-based dark matter detectors, especially DARWIN, could detect neutrinos from MeV-mass dark matter that annihilates to the third neutrino mass eigenstate, with projected sensitivity competitive with Super-Kamiokande.","lead":"The authors calculate how well large xenon dark matter detectors could see neutrinos from MeV-scale dark matter that annihilates almost entirely into the third neutrino mass eigenstate. They find DARWIN could probe this scenario in a region overlapping future JUNO and Hyper-Kamiokande sensitivities, enabling coincidence searches.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The DARWIN sensitivity limits in Fig. 1 rest on a printed test statistic (Eq. 5) that is not a valid Poisson likelihood and is internally inconsistent, so the central competitive-with-SK claim lacks quantitative support as written.","rationale":"Good-faith summary: the paper's central claim is that DARWIN, via CEνNS, can probe χχ→ν3ν3 in the roughly 10–100 MeV mass range at a level comparable to SK, with a discovery region overlapping JUNO and HK. The condition needed for this to hold is that the DARWIN sensitivity curve in Fig. 1 is computed correctly. The printed test statistic, Eq. (5), fails that condition: it is not a Poisson likelihood ratio, its 'observed' count depends on the fitted nuisance and signal parameters, and the Wilks threshold is chosen with the wrong number of degrees of freedom for a one-parameter upper limit. Without code, this is the weakest link in the quantitative argument. I do not treat the exclusive annihilation-to-ν3 scenario as a fatal flaw: it is explicitly a model assumption, motivated in the paper by majoron and scotogenic examples, and the DARWIN CEνNS channel is flavor-blind, so the sensitivity curve is not where that assumption breaks; the flavor assumption mainly sets the rescaling of the SK, JUNO, and HK comparison lines. Those rescalings are simple and plausible. The concern is therefore narrower: the main quantitative result is not reproducible from the text as it stands. Because the reader already reached CONDITIONAL and this concern supports that verdict, no change is needed; an independent correct-likelihood recomputation is the one concrete check that would settle it.","tokens_in":11243,"tokens_out":9235,"duration_ms":85573,"concrete_test":"Recompute the DARWIN 90% C.L. sensitivity curves for both exposure cases using the inputs stated in the paper — flux from Eq. (1) with J_av = 5, CEνNS cross section from Eq. (3), Helm form factor from Eq. (4), XENON1T efficiency and energy window from Ref. [73], the same solar and atmospheric neutrino spectra, and the same pull uncertainties — but replace Eq. (5) with the correct Asimov Poisson profile likelihood and a one-d.o.f. threshold. Then compare the new <sigma v> limits with the recast SK curve in Fig. 1. If the corrected DARWIN line shifts by more than ~30% in normalization, or its crossing with the recast SK limit moves by more than a factor ~1.5 in <sigma v>, the 'competitive with SK' claim is not yet established; if the shift is negligible, Eq. (5) is a harmless typo and the claim stands.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Eq. (5) is not a valid Poisson log-likelihood ratio. The prefactor contains N^i_sig + x N^i_sol + y N^i_atm + N^i_obs, and then N^i_obs is defined as N^i_sig + (1+x) N^i_sol + (1+y) N^i_atm; the 'observed' count is thus itself a function of the signal and pulled backgrounds. The logarithm uses the background-only expectation N^i_atm + N^i_sol in the numerator rather than the full model prediction. A correct Asimov Poisson profile likelihood for background-only pseudo-data n_i = N^i_sol + N^i_atm has -2 ln lambda = 2 sum_i [mu_i - n_i + n_i ln(n_i/mu_i)] + (x/sigma_sol)^2 + (y/sigma_atm)^2, with mu_i = N^i_sig + (1+x) N^i_sol + (1+y) N^i_atm. Eq. (5) matches neither this nor any standard limiting form. Additionally, a 90% C.L. upper limit on the single parameter <sigma v> should use a one-d.o.f. profile-likelihood threshold Delta chi^2 = 2.71, not the stated two-d.o.f. threshold 4.61; the latter weakens the limit by about a factor sqrt(4.61/2.71) ~ 1.3. The DARWIN curves in Fig. 1 are generated from Eq. (5), and the headline claim that DARWIN is competitive with SK is a comparison of these curves with the recast SK line. The statement that the DARWIN limits are consistent with Ref. [84] is helpful, but it cannot validate Eq. (5) as printed, and no code is provided.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies indirect detection of MeV-scale dark matter that annihilates exclusively to the third neutrino mass eigenstate, nu_3, whose electron-flavor content is only |U_e3|^2 ~ 0.02. Because conventional neutrino detectors are most sensitive to electron neutrinos, the expected signal in those detectors is strongly suppressed; the paper argues that large-scale xenon dark matter detectors, via coherent elastic neutrino-nucleus scattering, can probe this channel. Using a monochromatic Galactic neutrino flux with Jav = 5, the standard CE-NS cross section, and solar and atmospheric neutrino backgrounds, the authors compute expected 90% C.L. sensitivities for XENON1T and two DARWIN-like exposures and compare them with recast Super-K, Hyper-K, and JUNO limits. They conclude that DARWIN can be competitive with the current Super-K limits and that its projected sensitivity overlaps with JUNO and Hyper-K discovery regions for dark matter masses in the tens of MeV.","tokens_in":11654,"tokens_out":15072,"duration_ms":137320,"significance":"If the quantitative results survive a corrected statistical treatment, the paper makes a useful and falsifiable point: flavor-blind CE-NS detectors can probe neutrinophilic dark matter scenarios that are nearly invisible to electron-neutrino-based telescopes. The flux and rate formulas in Secs. II and III A are standard and transparent, the J-factor choice matches the previous SK/HK/JUNO analyses used for comparison, and the authors explicitly state several caveats, including the neglect of the extragalactic contribution and of the diffuse supernova neutrino background. The predicted overlap between DARWIN and JUNO/HK is a concrete experimental target. However, the central DARWIN sensitivity curves in Fig. 1 rest on a printed test statistic that is not a valid Poisson likelihood ratio and on an incorrect number of degrees of freedom, so the headline competitive-with-SK claim is not yet quantitatively established by the manuscript as written.","major_comments":[{"comment":"Equation (5) is not a valid Poisson log-likelihood-ratio statistic. For background-only Asimov data n_i = N^i_sol + N^i_atm and model expectation mu_i = N^i_sig + (1+x) N^i_sol + (1+y) N^i_atm, the Poisson profile likelihood ratio is -2 ln lambda = 2 sum_i [mu_i - n_i + n_i ln(n_i/mu_i)] + (x/sigma_sol)^2 + (y/sigma_atm)^2. As printed, Eq. (5) has prefactor N^i_sig + x N^i_sol + y N^i_atm + N^i_obs, which equals 2 N^i_sig + (1+2x) N^i_sol + (1+2y) N^i_atm, and uses log[(N^i_atm+N^i_sol)/N^i_obs]; this matches neither the correct Poisson term nor any standard limiting form. Because the DARWIN sensitivity curves in Fig. 1 are generated from Eq. (5), the central claim that DARWIN is competitive with Super-K is not quantitatively supported by the printed analysis. The statement that the limits are consistent with Ref. [84] is helpful, but it does not validate Eq. (5) as written.","section":"III B, Eq. (5)"},{"comment":"The stated 90% C.L. threshold is incorrect. After minimizing over the nuisance parameters x and y, the parameter of interest is <sigma v> alone, so under Wilks' theorem the appropriate one-dimensional profile-likelihood threshold is Delta chi^2 = 2.71, not the two-degree-of-freedom value Delta chi^2 = 4.61 used in the text. Using 4.61 weakens the derived cross-section limit by approximately sqrt(4.61/2.71) ~ 1.3. The DARWIN curves in Fig. 1 should be regenerated with the correct one-dimensional threshold.","section":"III B, text after Eq. (5)"}],"minor_comments":[{"comment":"The XENON1T red region is computed as a zero-background Poisson sensitivity (lambda ~ 2.3 events), not from the actual XENON1T event sample; please label it as an expected sensitivity or provide an observed limit, otherwise it may be read as an existing exclusion.","section":"III B, Fig. 1"},{"comment":"The recast of the SK/HK/JUNO limits assumes that the sensitivity of those detectors is dominated by the electron-neutrino component; please state this explicitly, since the factors 1/3 / |U_e3|^2 and 0.55 / |U_e3|^2 are otherwise not self-evident.","section":"I, recasting paragraph"},{"comment":"DUNE is mentioned in the abstract and conclusions as a potential coincident detector, but no DUNE sensitivity calculation appears in the paper; either add the estimate or soften the claim.","section":"I and IV"},{"comment":"The Heaviside function Theta(E_max^r - E_r) is redundant because the differential cross section in Eq. (3) already vanishes above E_max^r; this is harmless but should be noted or removed.","section":"II, Eq. (2)"},{"comment":"Please fix the inconsistent 'DAR WIN'/'DARWIN' spacing and the typographical errors 'scenrio' in the Introduction and 'DARIWN' in Sec. III B.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the paper is within the journal's scope and the underlying physics idea is sound, but the central figure is built on an invalid printed statistic and a wrong d.o.f. threshold, and no code is provided. I recommend major revision with a corrected Eq. (5), the correct one-dimensional threshold, a clearly labeled XENON1T sensitivity curve, and ideally a small validation table or code release for the DARWIN curves."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a clean, honest phenomenological paper about a niche but viable scenario — MeV-scale DM annihilating exclusively to the third neutrino mass eigenstate — and it makes the case that DARWIN-class detectors can compete with Super-K in this channel. The DARWIN sensitivity projection itself is not new: the authors state it is consistent with McKeen & Raj (Ref. [84]). What the paper adds is a clear side-by-side comparison of CEνNS detectors and classic neutrino detectors for this specific channel, and a sensible discussion of what a coincident signal would mean. That is useful for experimental planning.\n\nThe flux and CEνNS rate formulas are standard and transparent, and the XENON1T limit uses a simple, defensible Poisson counting argument. The authors are upfront about the model assumption — exclusive annihilation to ν3 — and about why it matters: the 2% |Ue3|² suppression is what makes the channel otherwise hard to see. I think that is a legitimate benchmark even though it is not derived here.\n\nThe soft spot is the DARWIN test statistic. Eq. (5) as printed is not a valid Poisson likelihood-ratio statistic. The 'observed' count is defined as signal plus pulled backgrounds, the log term uses the background-only expectation in a way that does not match an Asimov likelihood, and the stated 2 d.o.f. threshold for a 90% upper limit on a single parameter is not right; it should be 1 d.o.f., Δχ² = 2.71. Using 4.61 makes the limits conservative, but '90% C.L.' is then not accurate. Because the curves are consistent with Ref. [84], the numerical results may survive a proper recalculation, but as printed the central sensitivity curves are not quantitatively supported. This is fixable. No code is provided, which makes it harder to check, but the formulas are standard enough that a corrected version should be easy to verify.\n\nNet: this deserves a serious referee. The scenario is worth laying out for the DARWIN/JUNO/HK era, and the issues are exactly the kind a referee would catch. I would not build on the quantitative curves until the statistic is fixed and the threshold corrected.","headline":"Useful scenario paper on DM->nu3 detection via CEvNS; the DARWIN projection is not new and the printed test statistic needs fixing before the curves can be trusted.","tokens_in":12187,"tokens_out":5068,"would_cite":false,"duration_ms":44361,"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":"Planned xenon dark matter detectors could match Super-Kamiokande's reach for MeV-scale dark matter annihilating to the third neutrino mass eigenstate.","keywords":["neutrinophilic dark matter","MeV dark matter","third neutrino mass eigenstate","coherent elastic neutrino-nucleus scattering","CEvNS","direct dark matter detectors","neutrino telescopes","Galactic dark matter annihilation"],"falsifier":"A 200 tonne-year DARWIN-like exposure with a 1 keV nuclear-recoil threshold that finds no excess above solar and atmospheric neutrino backgrounds in the relevant recoil-energy bins, while JUNO or Hyper-Kamiokande measures an electron-neutrino flux from the Galactic center larger than the level predicted from $|U_{e3}|^2 \\simeq 0.02$, would exclude exclusive annihilation to the third mass eigenstate.","tokens_in":11072,"feed_emoji":"🌌","tokens_out":16021,"duration_ms":127280,"temperature":0.7,"pith_summary":"MeV-scale dark matter that annihilates only into neutrinos is one of the least constrained dark matter scenarios, because the final-state particles barely interact. This paper considers the flavor-hidden version of that scenario: dark matter of mass 10–100 MeV annihilating exclusively to the third neutrino mass eigenstate, whose electron-flavor fraction is only about $|U_{e3}|^2 \\approx 2\\%$. In that case the electron-neutrino flux is suppressed enough that conventional neutrino telescopes lose up to two orders of magnitude in sensitivity. The paper shows that large xenon-based dark matter detectors, which see neutrinos through flavor-blind coherent elastic neutrino-nucleus scattering, can recover much of that lost sensitivity, and that a planned detector like DARWIN would be competitive with Super-Kamiokande and overlap the projected reach of JUNO and Hyper-Kamiokande. A signal seen in both detector types would be a strong indication that dark matter annihilates to neutrino mass states rather than to charged particles.","feed_headline":"DARWIN-class xenon detectors could match Super-Kamiokande on MeV DM","feed_subtitle":"Nuclear recoils let xenon detectors catch a neutrino signal ordinary telescopes miss.","key_machinery":"The central mechanism is coherent elastic neutrino-nucleus scattering (CEvNS): a neutrino scatters off the whole xenon nucleus, producing a nuclear recoil with a cross section enhanced by the square of the weak charge, roughly the neutron number squared. Because CEvNS is flavor-blind, it registers the muon- and tau-dominated $\\nu_3$ flux that electron-flavor-sensitive detectors miss. The companion identity is the small electron-flavor content of the third mass eigenstate, $|U_{e3}|^2 \\simeq 0.02$, which keeps the electron-neutrino flux low in this scenario and is the reason the authors divide the recast neutrino-telescope limits by this factor. The calculation combines the delta-function Galactic flux at $E_\\nu=m_\\chi$, the CEvNS differential cross section with a standard nuclear form factor, and a 1 keV-binned likelihood that marginalizes over solar and atmospheric neutrino normalization uncertainties.","core_discovery":"Restricting the annihilation channel to $\\chi\\chi\\to\\nu_3\\bar\\nu_3$, the authors compute the 90% confidence sensitivity of xenon dark matter detectors to the resulting Galactic neutrino flux, modeled as a monochromatic line at $E_\\nu = m_\\chi$ with the canonical halo $J$-factor $J_{\\rm av}=5$. They recast the best existing limits, Super-Kamiokande plus projected JUNO and Hyper-Kamiokande sensitivities, by multiplying by $1/3$ and dividing by $|U_{e3}|^2$ to account for the small electron-flavor content of the third mass eigenstate. They find that XENON1T is not competitive, but a DARWIN-like detector, with either 40 tonne-years of exposure using XENON1T's efficiency and energy window or 200 tonne-years with 100% efficiency down to a 1 keV nuclear-recoil threshold, can probe annihilation cross sections in the $10^{-25}$–$10^{-24}~\\mathrm{cm^3\\,s^{-1}}$ range for masses around 25–100 MeV. That makes dark matter detectors competitive with the current Super-Kamiokande limits and places a non-excluded, discoverable region of parameter space within reach of both DARWIN-type CEvNS detectors and future neutrino telescopes. The paper's central claim is that planned large-scale dark matter detectors become neutrino telescopes for this specific flavor-hidden annihilation channel, and that coincident signals across direct-detection and neutrino experiments would indicate dark matter annihilation to neutrino mass eigenstates.","pith_inferences":["Because the paper notes that the extragalactic diffuse neutrino flux from annihilation could be comparable to the Galactic contribution, real DARWIN sensitivity might be somewhat better than the Galactic-only limits shown here.","The same CEvNS-based strategy should apply to any low-energy source of non-electron-flavor neutrinos, such as supernova neutrinos or decaying heavy neutrinos, making large dark matter detectors general-purpose neutrino observatories beyond the dark matter search.","A null result in DARWIN combined with an electron-neutrino excess in JUNO or Hyper-Kamiokande would not exclude all neutrinophilic dark matter, but it would rule out the exclusive-$\\nu_3$ branch and push models toward democratic or electron-flavor annihilation channels.","The recasting assumes vacuum oscillations with decoherence and negligible matter effects; a sharp measurement of the neutrino mass ordering and of the $\\theta_{23}$ octant would refine the predicted electron-flavor leakage and sharpen or weaken the reach estimate."],"forward_implications":["DARWIN-like exposures of 40–200 tonne-years can probe annihilation cross sections down to roughly $10^{-25}~\\mathrm{cm^3\\,s^{-1}}$ in the 25–100 MeV mass range, matching current Super-Kamiokande limits.","The projected sensitivity regions of DARWIN and of JUNO and Hyper-Kamiokande overlap, so the same model could be confirmed by two independent detector types.","A coincident signal in a CEvNS detector and a flavor-sensitive neutrino telescope would strongly indicate annihilation to neutrino mass states rather than to charged particles.","Solar neutrino backgrounds set a practical floor: dark matter masses below about 15 MeV are difficult to reach because solar-neutrino recoils dominate below roughly 3 keV.","A non-observation by XENON1T does not constrain this channel as strongly as Super-Kamiokande, but the larger planned exposures change that conclusion."],"supporting_citations":[{"why":"Supplies the canonical average J-factor $J_{\\rm av}=5$ and Galactic flux normalization used for the DM annihilation flux.","marker":"[16]"},{"why":"Provides the flux formula $dN_\\nu/dE_\\nu$ for dark matter annihilation that this work applies to $\\nu_3$.","marker":"[17]"},{"why":"Gives the Super-Kamiokande 90% C.L. limits that are recast to $\\nu_3$ and serve as the comparison benchmark.","marker":"[25]"},{"why":"Provides the Hyper-Kamiokande $\\nu_e$-channel sensitivity used for the recast with the $0.55/|U_{e3}|^2$ factor.","marker":"[28]"},{"why":"Gives the projected Hyper-Kamiokande sensitivity to DM annihilation to neutrinos used to define the overlapping discovery region.","marker":"[36]"},{"why":"Gives the JUNO projected sensitivity used in the multi-detector comparison.","marker":"[48]"},{"why":"Derives the coherent elastic neutrino-nucleus scattering cross section that produces the xenon recoil signal.","marker":"[64]"},{"why":"Reports the first observation of CEvNS, establishing the detection channel as experimentally viable.","marker":"[65]"},{"why":"Provides the earlier sensitivity calculation for dark matter annihilating to a tau-neutrino flux in large detectors, with which these results are compared.","marker":"[84]"}],"fun_headline_variants":["Xenon detectors can catch neutrinos from dark matter annihilation","Flavor-hidden neutrinos from DM become visible in xenon","MeV-scale neutrinophilic DM: xenon detectors join the hunt","Large xenon detectors rival neutrino telescopes for DM signal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands on the assumption that dark matter annihilates exclusively to the third neutrino mass eigenstate, whose electron-flavor content is only about 2%; if annihilation instead produced democratic flavor ratios or flavor states, neutrino oscillations would create a large electron-neutrino component and conventional detectors would regain sensitivity, changing the projected discovery region.","fun_headline_variants_meta":{"raw":{"variants":["Xenon detectors can catch neutrinos from dark matter annihilation","Flavor-hidden neutrinos from DM become visible in xenon","MeV-scale neutrinophilic DM: xenon detectors join the hunt","Large xenon detectors rival neutrino telescopes for DM signal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000301,"raw_usage":{"total_tokens":1791,"prompt_tokens":1055,"completion_tokens":736,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":671,"completion_tokens_details":{"reasoning_tokens":665}},"tokens_in":671,"tokens_out":736,"duration_ms":6496,"temperature":1.0,"reasoning_tokens":665,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T05:46:29.203215+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A 200 tonne-year DARWIN-like exposure with a 1 keV nuclear-recoil threshold that finds no excess above solar and atmospheric neutrino backgrounds in the relevant recoil-energy bins, while JUNO or Hyper-Kamiokande measures an electron-neutrino flux from the Galactic center larger than the level predicted from $|U_{e3}|^2 \\simeq 0.02$, would exclude exclusive annihilation to the third mass eigenstate.","supporting_citations":[],"review_version":1}