{"id":"3015ec77-247e-430d-861d-ad3ea49e2ff0","arxiv_id":"2511.11856","paper_version":3,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"SNO+ measures Δm²₂₁ = (7.93^{+0.21}_{-0.24})×10⁻⁵ eV² from 1.46 ktonne-years of reactor antineutrino data and detects geoneutrinos at 49^{+13}_{-12} TNU.","lead":"SNO+ measured reactor antineutrinos over three years and extracted the neutrino mass-squared difference Δm²₂₁ = (7.93^{+0.21}_{-0.24})×10⁻⁵ eV². This is the second long-baseline reactor measurement of this parameter, approaching KamLAND's precision, and it also detects geoneutrinos at 4.1σ significance.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Primary fit floats only (α,n) normalizations; an energy-correlated shape error in the dominant multiple-proton-scattering background could bias Δm²21.","rationale":"The reader's weakest assumption identified exactly this: only a floating normalization for the (α,n) background, with no explicit energy-correlated shape uncertainty, could affect the fitted Δm²21. The central claim is the measurement of Δm²21 from the spectral fit, and the (α,n) background dominates at low energies where the oscillation pattern is rapid. The paper does not include a shape systematic for this component in the primary fit, and the AmBe data—which could directly validate the shape—are used only for the classifier. The difference between the primary fit and the classifier-cut fit (Δm²21 = 7.93 vs 7.56) suggests the result is sensitive to the modeling of this background. Therefore, this is a load-bearing concern, and the paper should be accepted only after the authors demonstrate, using the AmBe data or an independent shape parameterization, that energy-dependent shape uncertainties do not significantly shift the headline value. The proposed concrete test is feasible with the already-deployed AmBe source and would settle the concern quantitatively.","tokens_in":11554,"tokens_out":8786,"duration_ms":73720,"concrete_test":"Use the July 2025 AmBe dataset—which mimics (α,n) proton scatters—to measure the prompt spectral shape of multiple-proton-scattering events and compare it with the simulated spectrum used in the (α,n) background model. Then re-run the primary spectral fit with an energy-dependent scaling of the (α,n) multiple-scattering component (e.g., a linear slope or binned ratios) constrained by the AmBe data, and observe the shift in the best-fit Δm²21. If the shift exceeds ~0.05×10⁻⁵ eV², the current normalization-only treatment is insufficient and the shape uncertainty must be included in the quoted systematic error.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is the spectral measurement of Δm²21 from reactor antineutrinos. The dominant background below 3.5 MeV is the (α,n) multiple-proton-scattering continuum, whose spectral shape is taken from simulation. In the 'Spectral Analysis and Results' section, the fit only floats 'scalings to the cross section and branching ratios' for the (α,n) channels—i.e., normalizations of subcomponents. No energy-correlated shape uncertainty is included. The AmBe calibration is used to assess the classifier, but not to validate the spectral shape of the (α,n) prompt continuum. If the true (α,n) spectrum has an energy-dependent error (e.g., a slope or wiggles in the 1–4 MeV range), the fitted oscillation pattern—which is an oscillatory function of L/E across the full energy range—could be partially absorbed or distorted, shifting Δm²21. The quoted uncertainty on Δm²21 is +0.21/−0.24×10⁻⁵ eV², and the paper states the measurement is statistically dominated; a shape bias of only a few percent could produce a shift of this order. The classifier-cut fit yields Δm²21 = (7.56±0.17)×10⁻⁵ eV², lower than the primary fit by ~0.4×10⁻⁵, hinting at sensitivity to the (α,n) shape. The absence of an explicit shape systematic in the primary fit is a genuine soft spot.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The SNO+ Collaboration reports a spectral analysis of reactor antineutrinos using 1.46 ktonne-years of data collected from May 2022 through July 2025. An unbinned extended-likelihood fit to the prompt-energy spectrum yields Δm²₂₁ = (7.93^{+0.21}_{-0.24})×10⁻⁵ eV² and sin²θ₁₂ = 0.505±0.134, with a geoneutrino signal of 49^{+13}_{-12} TNU (4.1σ) obtained after applying a new (α,n) event classifier. Combining the SNO+ data with PDG 2025 global constraints, the authors report updated world averages Δm²₂₁ = (7.63±0.17)×10⁻⁵ eV² and sin²θ₁₂ = 0.310±0.012. The paper includes two independent fitters, a comprehensive set of detector/flux/background systematics, and a first AmBe-based calibration of the (α,n) classifier.","tokens_in":11925,"tokens_out":6594,"duration_ms":54430,"significance":"If the result holds, SNO+ becomes the second experiment to measure reactor antineutrino oscillations at solar-scale baselines, with a precision approaching KamLAND's. The distinct baseline distribution (240, 350, and 355 km) provides an independent cross-check of the KamLAND measurement and a complementary probe of the solar mass-splitting parameter. The paper is also notable for the first application of an (α,n) pulse-shape classifier in a large liquid-scintillator detector, which improves the geoneutrino measurement. Strengths include the use of two independent fitters with consistent results, explicit propagation of detector, reactor-flux, and background systematics, and a calibration deployment (AmBe) used to quantify classifier-related uncertainties. The main weaknesses are the absence of an energy-correlated shape uncertainty for the dominant (α,n) background in the spectral fit and a potential double-counting of SNO+ data in the PDG-based global combination; both are addressable.","major_comments":[{"comment":"The dominant background below ~3.5 MeV is the multiple-proton-scattering continuum from (α,n). In the fit description, only scalar 'scalings to the cross section and branching ratios' are floated for the (α,n) channels; no energy-correlated shape nuisance is included. Because Δm²₂₁ is extracted from the oscillatory spectral distortion across the full 0.9–8 MeV range, an energy-dependent error in the simulated (α,n) spectrum (e.g., a slope or resonance-like wiggle from the ¹³C(α,n) cross section) could bias Δm²₂₁ by an amount comparable to the quoted statistical uncertainty of ≈0.2×10⁻⁵ eV². The AmBe source has a different neutron spectrum (⁹Be(α,n)¹²C), so it does not directly validate the ¹³C(α,n) spectral shape. The third fit in Table III, which suppresses the (α,n) background, gives Δm²₂₁ = (7.56±0.17)×10⁻⁵ eV², a ~0.4×10⁻⁵ shift relative to the unconstrained fit, indicating some sens","section":"(α,n) Background / Spectral Analysis and Results"},{"comment":"The global combination constrains the SNO+ fit with Gaussian constraints to the PDG 2025 values (Ref. [16]) and reports updated world averages Δm²₂₁ = (7.63±0.17)×10⁻⁵ eV² and sin²θ₁₂ = 0.310±0.012. However, the current SNO+ dataset includes the previously published SNO+ dataset (Ref. [2]) and is five times larger. If the PDG 2025 update already incorporates that earlier SNO+ result, then this combination double-counts SNO+ data, and the claimed impact ('a moderate increase... from (7.53±0.18) to (7.63±0.17)') is not a clean measure of the new data's constraining power. Please state explicitly whether the PDG constraints exclude the earlier SNO+ measurement; if they do not, recompute the combined fit using external constraints that do not contain SNO+ data, or provide a proper covariance-level combination.","section":"Spectral Analysis and Results / Summary"},{"comment":"The 4.1σ geoneutrino significance and the 49 TNU central value (Summary and Table III) rely on the classifier fit, where the dataset I classifier systematic is assumed to have the same energy dependence and magnitude as dataset II, with no direct AmBe deployment for dataset I. This assumption is acknowledged, but no cross-check is shown. Since the classifier cut substantially changes the background decomposition (Table II: (α,n) p-scatters drop from 63 to 22 counts), the geoneutrino error budget depends sensitively on this assumed systematic. Please perform a sensitivity test in which the dataset I classifier uncertainty is taken as fully correlated with dataset II (or a more conservative common uncertainty), and confirm that the 4.1σ significance is stable.","section":"Calibrations / Spectral Analysis and Results, Table III"}],"minor_comments":[{"comment":"The text refers to a 'Fischer discriminant'; the standard spelling is 'Fisher discriminant'.","section":"Calibrations"},{"comment":"The notation 'β's' and 'α's' is awkward; use 'β particles'/'α particles' or 'β decays' for clarity.","section":"General"},{"comment":"The contour labels '1, 2, 3' in the figure are not defined in the caption; please state that they correspond to confidence levels (e.g., 1σ, 2σ, 3σ).","section":"Figure 3"},{"comment":"Ref. [16] is cited as 'Phys. Rev. D110, 030001 (2024), and 2025 update.' If a specific 2025 PDG update reference exists, it should be cited directly; the current form is ambiguous about the exact update date and content.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a strong experimental result, and the standalone Δm²₂₁ measurement appears defensible modulo the (α,n) shape-systematic concern. The PDG double-counting issue may be resolvable by simply checking whether PDG 2025 already includes the earlier SNO+ paper; if not, that major comment reduces to a clarification request. The geoneutrino significance test is also straightforward. I recommend major revision rather than rejection because the central claim is sound in principle and the issues are local, fixable, and do not require a change in the scope of the analysis."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this paper deserves peer review and will likely be cited widely. It is an honest, careful measurement. The genuinely new things are the 5× dataset, the first application of the (α,n) classifier, and a geoneutrino signal at 4.1σ from a new site. The central claim — an independent long-baseline reactor measurement of Δm²₂₁ = (7.93+0.21−0.24)×10⁻⁵ eV² — is well-supported by the fit. Two independent fitters agree; the systematic treatment is unusually thorough for a letter; reactor flux uncertainties are tracked daily and cross-checked; the AmBe calibration is a nice first step.\n\nThe soft spot is real but minor: the dominant (α,n) multiple-proton-scattering background is taken from simulation, and the primary fit floats only normalizations, not an energy-correlated shape. The paper acknowledges this and provides a cross-check via the classifier cut, which gives 7.56±0.17 — about 1.3σ lower. That shift doesn't upend the result, but it suggests the stated statistical errors might be a bit optimistic at the low end. The same caveat applies to the geoneutrino rate, which is anti-correlated with the (α,n) normalization.\n\nThe global combination with PDG is standard; the self-citation overlap is negligible. The paper is transparent about classifier uncertainties, including the extrapolation from dataset II to dataset I.\n\nThis is a paper for neutrino physicists and geophysicists. It deserves serious refereeing; the referee should press on the (α,n) shape uncertainty and the 3% proton energy-scale uncertainty, but neither is load-bearing. I would accept for review.","headline":"A careful, statistically-dominated SNO+ measurement gives a second long-baseline reactor Δm²₂₁ close to KamLAND precision, with a real but non-fatal soft spot in the (α,n) background shape uncertainty.","tokens_in":12988,"tokens_out":2027,"would_cite":true,"duration_ms":18130,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Using 1.46 ktonne-years of reactor-antineutrino data, this paper reports a spectral fit to the solar-scale mass-squared difference yielding Δm²₂₁=(7.93^{+0.21}_{-0.24})×10⁻⁵ eV².","keywords":["reactor antineutrino oscillations","Δm²₂₁","solar neutrino mass splitting","geoneutrinos","inverse beta decay","liquid scintillator detector","(α,n) background","Fisher discriminant"],"falsifier":"Compare the reconstructed energy spectrum of neutron recoils from a deployed AmBe source in the SNO+ detector with the simulation used for the (α,n) background; any energy-correlated discrepancy exceeding the assigned systematic would invalidate the central value. Alternatively, a future long-baseline reactor experiment with different backgrounds and better energy resolution could independently verify Δm²₂₁ at comparable precision.","tokens_in":11461,"feed_emoji":"⚛️","tokens_out":10478,"duration_ms":75339,"temperature":0.7,"pith_summary":"Using 1.46 ktonne-years of data, this paper extracts the solar-scale neutrino mass-squared difference from the energy spectrum of reactor antineutrinos arriving at three baselines of 240, 350, and 355 km. The unbinned spectral fit yields Δm²₂₁=(7.93^{+0.21}_{-0.24})×10⁻⁵ eV², a precision that approaches the only earlier long-baseline reactor antineutrino measurement. The same dataset detects geoneutrinos at 49^{+13}_{-12} TNU, a 4.1σ signal, thanks to a new event classifier that suppresses the dominant (α,n) background. Combined fits with other neutrino data give a global Δm²₂₁=(7.63±0.17)×10⁻⁵ eV² and sin²θ₁₂=0.310±0.012, and the paper argues this independent reactor measurement sharpens the test of the three-flavor oscillation picture.","feed_headline":"Reactor antineutrinos set neutrino mass split at 7.93×10⁻⁵ eV²","feed_subtitle":"A 1.46 ktonne-year run at 240–355 km baselines gives a second independent measurement of the neutrino mass splitting.","key_machinery":"The central machinery is the reactor antineutrino survival probability P_ee, which produces an energy-dependent deficit dominated by sin²(Δm²₂₁ L/4E). The signal is detected via inverse beta decay: a prompt positron plus a delayed 2.2-MeV neutron-capture gamma, selected by a likelihood ratio on time, distance, and delayed energy. To handle the dominant background, the paper introduces a Fisher discriminant event classifier built from time-of-flight-corrected PMT hit times and radial position; it separates multiple-proton-scattering (α,n) signals from IBD positrons below 3.5 MeV, rejecting roughly 82–92% of the (α,n) events while retaining about 90% of geoneutrino and 60% of reactor IBDs. An","core_discovery":"The paper establishes that a spectral analysis of inverse-beta-decay events, selected by their prompt-positron and delayed-neutron-capture coincidence, can resolve the oscillation dip imposed by Δm²₂₁ on reactor antineutrinos traveling 240–355 km. The best fit to the two unconstrained oscillation parameters is Δm²₂₁=(7.93^{+0.21}_{-0.24})×10⁻⁵ eV² and sin²θ₁₂=0.505±0.134; constraining the mixing angle to the global value returns Δm²₂₁=(7.90^{+0.26}_{-0.31})×10⁻⁵ eV². The measurement is compatible with the global fit at 1.4σ. A separate fit that applies the (α,n) classifier reduces the dominant background by a factor of ~4–5 and yields a geoneutrino rate of 49^{+13}_{-12} TNU, a 4.1σ detectio","pith_inferences":["If the upward-shifted Δm²₂₁ persists as statistics accumulate, the 1.6σ discrepancy with solar neutrino fits could harden into a real tension, pointing either to underestimated reactor flux shape uncertainties or to new physics in the neutrino sector.","The (α,n) classifier is transferable in concept to other organic-scintillator neutrino detectors, where the same multiple-proton-scatter background limits the low-energy region; applying it elsewhere would provide a cross-check of this background model.","A dedicated AmBe-source measurement spanning the full prompt-energy range (0.9–4 MeV) would directly test the assumed (α,n) spectral shape; if the shape deviates, both the quoted Δm²₂₁ and the geoneutrino rate are impacted."],"forward_implications":["A second experiment now measures Δm²₂₁ with long-baseline reactor antineutrinos, providing an independent test of the value obtained by the first.","Combining the SNO+ spectral data with solar and other reactor results shifts the global Δm²₂₁ from (7.53±0.18) to (7.63±0.17)×10⁻⁵ eV², a moderate upward movement.","The geoneutrino measurement at 49 TNU with ~26% uncertainty is the third such detection and can be combined with other sites to constrain continental crust contributions and mantle heat production.","Further data and improved knowledge of the (α,n) classifier's systematic uncertainty are expected to reduce the statistical and systematic errors on both Δm²₂₁ and the geoneutrino rate."],"fun_headline_variants":["SNO+ reactor data pins neutrino mass split to 7.93×10⁻⁵ eV²","SNO+ independently confirms neutrino mass splitting: 7.93×10⁻⁵ eV²","SNO+ matches KamLAND precision with 7.93×10⁻⁵ eV² neutrino mass split","SNO+ delivers reactor neutrino mass splitting at 7.93×10⁻⁵ eV²"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The simulated energy shape of the dominant (α,n) multiple-proton-scattering background is taken to be correct up to a floating normalization; if its true spectrum has an energy-dependent shape error in the 1–4 MeV region, the fitted Δm²₂₁ could shift outside the quoted uncertainty.","fun_headline_variants_meta":{"raw":{"variants":["SNO+ reactor data pins neutrino mass split to 7.93×10⁻⁵ eV²","SNO+ independently confirms neutrino mass splitting: 7.93×10⁻⁵ eV²","SNO+ matches KamLAND precision with 7.93×10⁻⁵ eV² neutrino mass split","SNO+ delivers reactor neutrino mass splitting at 7.93×10⁻⁵ eV²"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001052,"raw_usage":{"total_tokens":4288,"prompt_tokens":810,"completion_tokens":3478,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":554,"completion_tokens_details":{"reasoning_tokens":3372}},"tokens_in":554,"tokens_out":3478,"duration_ms":24516,"temperature":1.0,"reasoning_tokens":3372,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T22:08:24.026490+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the reconstructed energy spectrum of neutron recoils from a deployed AmBe source in the SNO+ detector with the simulation used for the (α,n) background; any energy-correlated discrepancy exceeding the assigned systematic would invalidate the central value. Alternatively, a future long-baseline reactor experiment with different backgrounds and better energy resolution could independently verify Δm²₂₁ at comparable precision.","supporting_citations":[],"review_version":1}