{"id":"271906ad-c9cf-4e98-9f0f-c08aa0ceea85","arxiv_id":"2608.12293","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"NOvA measures the antineutrino-hydrogen charged-current quasi-elastic cross section as 0.538 +/- 0.055 (flux) x 10^-38 cm^2 at 1.9 GeV, the most precise such measurement.","lead":"The NOvA collaboration reports the most precise measurement of the total cross section for antineutrinos scattering off hydrogen, with 35,509 signal events. The result is precise enough that its non-flux uncertainty is a quarter of the beam flux uncertainty, which could help calibrate future neutrino beams.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"CR/SR transfer of the 30% kHitID correction is asserted but not quantitatively tested; a data-side composition mismatch would bias the dominant background subtraction beyond the quoted background systematic.","rationale":"The reader's weakest assumption correctly identifies the CR/SR transfer as the soft spot. The paper provides substantial support for the analysis: a blinded signal region, data-driven corrections, an internally consistent uncertainty decomposition, and a fake-data closure test that demonstrates robustness to parameter variations within the adopted model. However, that closure test does not exercise the specific assumption that the CR and SR background compositions match in data, and the dominant background correction (kHitID=1.30) is large enough that a modest composition mismatch would shift the cross section by several percent. The difference between the contained and uncontained central values (0.565 vs 0.515) reinforces the need for a quantitative transfer check. Because the analysis is otherwise well-executed and the concern is testable with the released data, the appropriate outcome is conditional acceptance: the transfer assumption should be validated quantitatively before the quoted precision is taken at face value. If the proposed test shows the k factors are stable across the CR/SR boundary, the ACCEPT verdict would stand.","tokens_in":12978,"tokens_out":14065,"duration_ms":138584,"concrete_test":"Using the data release, recompute kHitID separately in two disjoint sub-regions of the HitID control region that bracket the signal region in KineID (one adjacent to the SR, one farther away), and compare the two estimates. If they differ by more than the control-region statistical uncertainty, or if applying the average kHitID to the SR yields a background prediction that fails to match data in an independent sideband variable such as reconstructed muon angle, then the single transferred correction is not valid and the quoted background systematic is underestimated.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The most load-bearing assumption is that the data-driven correction factors measured in the control regions transfer exactly to the signal region. In the contained sample, backgrounds outnumber signal by 27,839 to 15,103 (purity 41%), and kHitID=1.30 means a 30% upward correction to the MC background efficiency. The only stated justification is that the CR and SR background compositions are 'similar' (Event Selection section; Fig. S3), with no quantitative similarity metric or uncertainty attached. The fake-data studies (Fig. 3, S6) vary the underlying MC parameters but generate both CR and SR from the same model, so the correction factors absorb those variations by construction; they cannot detect a data-MC mismatch in the relative CCQE/MEC/RES fractions between CR and SR, nor a mismatch in the correlation between HitID and KineID. Because backgrounds are about 1.8 times the signal, a 2% error in the transferred k factor becomes about 3.7% of the signal yield and hence of the cross section, comparable to the 10.2% flux uncertainty and much larger than the quoted 1.2% background systematic. The 9% difference between the contained (0.565) and uncontained (0.515) results, while not statistically significant, is consistent with a composition-dependent transfer bias.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a measurement of the total cross section for the exclusive charged-current quasi-elastic process antineutrino-muon on hydrogen, \\bar\\nu_\\mu H -> \\mu^+ n, using the NOvA near detector with a 1.2x10^21 proton-on-target exposure. Events are selected in two mutually exclusive samples (contained and uncontained muons) using two BDT discriminants, HitID and KineID, with a blinded signal region and control regions used to derive data-driven corrections to simulated backgrounds. After background subtraction, the combined sample yields 35,509 signal events, and the cross section is extracted from a simple counting formula. The final flux-averaged result is 0.538 +/- 0.009 (stat) +/- 0.010 (syst) +/- 0.055 (flux) x 10^-38 cm^2 at an average energy of 1.9 GeV, claimed as the most precise measurement of this process to date.","tokens_in":13207,"tokens_out":8390,"duration_ms":75580,"significance":"This measurement is scientifically valuable: it provides the highest-statistics sample of (anti)neutrino-hydrogen interactions measured to date and a nuclear-model-free cross-section benchmark at 1.9 GeV, complementing the existing BNL and MINERvA measurements. The authors have blinded the signal region, repeated the full analysis for each systematic variation, and provided a data release, all of which strengthen reproducibility. The main limitation in the current manuscript is the unquantified transfer of data-driven correction factors from control regions to the signal region; because the signal purity is modest (41% contained, 36% uncontained), the precision of the final result rests on this assumption.","major_comments":[{"comment":"The assumption that the control-region corrections transfer to the signal region is not quantitatively validated. For the contained sample, k_HitID = 1.30 and backgrounds outnumber signal by 27,839 to 15,103; a 2% error in the transferred correction factor would shift the subtracted background by about 3.7% of the signal, which is comparable to the 10.2% flux uncertainty and far larger than the quoted 1.2% background systematic. The fake-data tests in Fig. 3 and Fig. S6 generate both control and signal regions from the same simulation, so they verify compensation of model variations but cannot detect a data-MC mismatch in the relative background composition between the regions. Please add a quantitative similarity metric (e.g., a composition-weighted closure test), vary the control-region definitions, or assign an explicit systematic uncertainty to the transfer.","section":"Data Corrections; Eq. (4); Table II"},{"comment":"The momentum reconstruction for the uncontained sample is not described. The text quotes a 3.5% range-based momentum resolution only for tracks stopping in the detector, while uncontained muons exit the detector; nevertheless the uncontained sample provides 20,670 of the 35,509 signal events and KineID and the calculated neutron kinematics depend on the muon momentum p_mu. Please specify how the muon momentum and its resolution are obtained for exiting tracks, including any use of the downstream muon detector, and how the associated uncertainties enter Table II.","section":"Event Selection; Eqs. (1)-(3)"},{"comment":"The k_HitID and k_KineID corrections are applied only to background efficiencies, while the signal efficiency epsilon_QEH in Eq. (5) is taken from Monte Carlo without a data-driven correction. Because epsilon_QEH is only 3.6% (contained) and 4.8% (uncontained), a small data-MC difference in the BDT response for signal events propagates directly into the measured cross section. The quoted QEH modeling uncertainty and the fake-data tests do not directly constrain this data-MC difference; please provide a validation of the signal efficiency using a signal-enriched sideband or add an explicit systematic uncertainty.","section":"Data Corrections; Eq. (5)"}],"minor_comments":[{"comment":"In the supplemental text, the reference to 'Figure S6' for the efficiency and purity plots appears to be a typo; the corresponding caption is FIG. S5, while FIG. S6 shows the correlation plot.","section":"Supplemental Material"},{"comment":"Reference [62], cited for the best linear unbiased estimator, is unusual (D. S. Huang, Regression and econometric methods, Wiley, 1930) and should be replaced with a standard citation for BLUE or Gauss-Markov estimation.","section":"References"},{"comment":"The acronyms CCQE and QEH are both used throughout; please define QEH at first use and consistently distinguish it from charged-current quasi-elastic interactions on nuclear targets.","section":"Event Selection"}],"recommendation":"major_revision","confidential_remarks":"The reader's report recommended accept, but the control-region to signal-region transfer assumption is load-bearing and not quantitatively tested. I recommend asking the collaboration for a closure test or an explicit transfer systematic before publication. The measurement itself is valuable and the requested additions are feasible within the paper's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline is simple: this is the first precise total cross section for \\bar\\nu_\\mu H -> \\mu^+ n, with 35,509 signal events versus the 13 from BNL. The extraction is straightforward counting, the signal region was blinded, and the systematic uncertainties are propagated by rerunning the full analysis for each variation. That is real care, and the paper is a genuine advance in a subfield where measurements on free protons are scarce.\n\nWhat is new here is the measurement itself, not the method. The selection follows the DUNE-proposal paper by Duyang et al., and the paper cites that work properly. The result -- 0.538 +/- 0.009 +/- 0.010 +/- 0.055 x 10^-38 cm^2 at 1.9 GeV -- is about four times more precise than the flux uncertainty, which makes it useful for flux calibration, though not field-shaping.\n\nThe soft spot is the control-region (CR) to signal-region (SR) transfer. The fractional correction k_HitID is 1.30 for the contained sample, meaning the MC background efficiency is scaled up by 30%. The paper says the CR and SR background compositions are \"similar\" and shows them in Fig. S3, but I do not see a quantitative similarity test or an uncertainty attached to the transfer itself. The 10,000 fake-data experiments vary model parameters, but they generate both CR and SR from the same underlying model, so they test the compensation mechanism, not a data-side composition mismatch. Given that backgrounds outnumber signal by about 1.8 to 1 in the contained sample, a small error in the transferred k factor becomes a few percent on the cross section -- bigger than the quoted 1.2% background systematic. The 9% difference between the contained and uncontained results is not statistically significant, but it is in the direction one might expect from a composition-dependent bias. I would want the collaboration to either demonstrate quantitatively that the compositions are matched or add a systematic for the residual difference.\n\nThat said, this is a caveat, not a fatal flaw. The measurement is defined directly from event counts, not from the GENIE cross-section input, so there is no circularity concern. The self-citation in Ref. [9] is appropriate. The paper deserves serious refereeing, and I would send it back to the collaboration asking for an explicit CR/SR transfer test -- but I would be surprised if that request leads to a change in the central value. If I worked on neutrino cross sections, I would cite this paper within the year.","headline":"First genuinely precise total cross section for antineutrino-hydrogen CCQE; a solid measurement whose main soft spot is the untested transfer of the 30% control-region background correction.","tokens_in":13777,"tokens_out":1930,"would_cite":true,"duration_ms":19180,"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":"Muon antineutrinos scattering on free protons in the NOvA near detector give a total charged-current quasi-elastic cross section of $0.538 \\pm 0.009 \\pm 0.010 \\pm 0.055 \\times 10^{-38}$ cm$^2$ at 1.9 GeV, the most precise measurement of…","keywords":["antineutrino charged-current quasi-elastic scattering","hydrogen target cross section","NOvA near detector","NuMI antineutrino beam","neutron detection","control region data corrections","neutrino flux uncertainty"],"falsifier":"Move the signal-region boundaries in the HitID-KineID plane and check that the extracted cross section stays constant within quoted uncertainties; a significant drift would indicate that the control-region corrections do not transfer, and the background composition difference between the regions could be tested directly against an independent high-statistics simulation.","tokens_in":12721,"feed_emoji":"🎯","tokens_out":11541,"duration_ms":91455,"temperature":0.7,"pith_summary":"The paper reports the most precise measurement to date of the total rate for muon antineutrinos scattering off free protons to produce a muon and a neutron, at a flux-averaged energy of 1.9 GeV. Using 35,509 signal events identified after background subtraction in the hydrogen-rich material of the NOvA near detector, the collaboration extracts a cross section of $0.538 \\pm 0.009 \\pm 0.010 \\pm 0.055 \\times 10^{-38}$ cm$^2$. Because the target is a single free proton, the result is a nuclear-model-free benchmark for antineutrino-nucleus calculations. Its 2.5% combined statistical and non-flux systematic uncertainty is more than four times smaller than the flux uncertainty, which is why the paper argues the measurement can be used to constrain the absolute muon-antineutrino flux in future analyses.","feed_headline":"35,509 hydrogen events set a new antineutrino cross-section benchmark","feed_subtitle":"The 2.5% precision on free-proton scattering can tighten beam-flux estimates in neutrino experiments.","key_machinery":"The analysis is carried by a signal-region/control-region design built on two boosted-decision-tree discriminants. HitID encodes topological information about hit patterns and energy deposits near the vertex and around the neutron candidate, while KineID encodes global event kinematics derived from the assumed two-body final state, including transverse momentum balance and the opening angle between the muon and the neutron. Control regions with background compositions similar to the signal region are used to measure in data the efficiencies of the sequential cuts, and correction factors $\\kappa = \\varepsilon_{Data}^{CR}/\\varepsilon_{MC}^{CR}$ are applied to the simulated background efficiencies in Eq. (4). These correction factors are anticorrelated with Monte Carlo variations, so that systematic shifts of up to about 60% in the simulation are compensated and the background systematic uncertainty is reduced by roughly an order of magnitude.","core_discovery":"The central claim is that the NOvA near detector has observed $\\bar\\nu_\\mu$-hydrogen charged-current quasi-elastic scattering, $\\bar\\nu_\\mu \\mathrm{H} \\to \\mu^+ n$, in a $1.2 \\times 10^{21}$ proton-on-target exposure and measured its total cross section with the highest statistics and best precision achieved for this process. After background subtraction, 35,509 signal events remain, and the flux-averaged cross section at 1.9 GeV is $\\sigma(\\bar\\nu_\\mu \\mathrm{H} \\to \\mu^+ n) = 0.538 \\pm 0.009 \\pm 0.010 \\pm 0.055 \\times 10^{-38}$ cm$^2$, where the uncertainties are statistical, non-flux systematic, and flux, respectively. The combined statistical and non-flux systematic uncertainty of 2.5% is more than four times smaller than the 10.2% integrated-flux uncertainty, so the measurement is limited by knowledge of the beam rather than by the extraction of the signal.","pith_inferences":["A direct extension would be to apply the same selection to neutrino-beam-mode data and extract the $\\nu_\\mu$ hydrogen cross section; the ratio of the two measurements would test the $q^2$ dependence of the axial form factor in a way that largely cancels flux uncertainty.","If the control-region similarity assumption holds, a larger exposure could push the non-flux uncertainty below 2%, making the hydrogen cross section one of the sharpest handles on the neutrino energy scale in long-baseline detectors.","The result implicitly sharpens nuclear-model constraints: any generator that matches this free-proton point while mismodeling events on carbon and chlorine is forced to attribute the difference to nuclear effects, separating free-nucleon from in-medium physics.","A natural cross-check is to move the signal-region boundaries in the HitID-KineID plane and confirm the extracted cross section is stable; a drift would indicate the background corrections do not transfer from control to signal regions."],"forward_implications":["The new cross-section value replaces a 13-event bubble-chamber measurement as the reference point for $\\bar\\nu_\\mu$-hydrogen charged-current quasi-elastic scattering near 2 GeV.","The 2.5% combined statistical and non-flux systematic precision makes the measurement a usable absolute-flux constraint for future neutrino oscillation analyses, as the paper expressly states.","The demonstrated stability of the data corrections against large Monte Carlo variations supports applying the same control-region technique to other channels and to detectors with composite targets.","Comparison of the measured value with generator predictions based on the Llewellyn-Smith model with a $z$-expansion axial form factor tests the modeling of the free-nucleon axial current at this energy."],"supporting_citations":[{"why":"Provides the only prior total antineutrino-hydrogen QEH cross-section measurement, the comparison anchor for the new result.","marker":"[12]"},{"why":"Defines the topological and kinematic selection of QEH events from CH2 targets that the analysis implements.","marker":"[9]"},{"why":"Supplies the PPFX hadron-production constraints used to predict the NuMI flux, the source of the dominant 10.2% flux uncertainty.","marker":"[17]"},{"why":"Provides the GENIE v3.0.6 event generator used for signal and background simulation.","marker":"[37]"},{"why":"Supplies the Llewellyn-Smith model used to simulate the QEH signal process.","marker":"[38]"},{"why":"Replaces the dipole axial form factor with the z-expansion model in the signal simulation.","marker":"[39]"},{"why":"Gives the inclusive charged-current data used to reweight the MEC and FSI models in the simulation.","marker":"[45]"},{"why":"Provides the NOMAD method of data-driven background efficiency corrections adopted in Eq. (4).","marker":"[58]"}],"fun_headline_variants":["Most precise antineutrino-hydrogen cross section from 35,509 events","NOvA pins down antineutrino-hydrogen scattering with 2.5% precision","35,509 events give cleanest antineutrino-proton cross section","Antineutrino-hydrogen cross section measured to 2.5% by NOvA","Most precise measurement of antineutrino scattering on hydrogen"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each control region has a background composition similar enough to the signal region that efficiency corrections measured there apply to the signal background; if the compositions differ, the corrections (notably $k_{\\rm HitID} = 1.30$ for contained events) would bias the subtracted background and shift the cross section.","fun_headline_variants_meta":{"raw":{"variants":["Most precise antineutrino-hydrogen cross section from 35,509 events","NOvA pins down antineutrino-hydrogen scattering with 2.5% precision","35,509 events give cleanest antineutrino-proton cross section","Antineutrino-hydrogen cross section measured to 2.5% by NOvA","Most precise measurement of antineutrino scattering on hydrogen"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000622,"raw_usage":{"total_tokens":2930,"prompt_tokens":1044,"completion_tokens":1886,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":660,"completion_tokens_details":{"reasoning_tokens":1780}},"tokens_in":660,"tokens_out":1886,"duration_ms":10837,"temperature":1.0,"reasoning_tokens":1780,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:10:08.651783+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Move the signal-region boundaries in the HitID-KineID plane and check that the extracted cross section stays constant within quoted uncertainties; a significant drift would indicate that the control-region corrections do not transfer, and the background composition difference between the regions could be tested directly against an independent high-statistics simulation.","supporting_citations":[{"cited_title":"Fanourakis, L","cited_arxiv_id":null,"evidence_quote":"Provides the only prior total antineutrino-hydrogen QEH cross-section measurement, the comparison anchor for the new result."}],"review_version":1}