{"id":"dd22c1c7-7426-42e8-9605-6140cdf22cb9","arxiv_id":"2412.04033","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"NA62 reports one candidate for a neutrino produced in a K+ to mu+ nu decay and detected in its liquid krypton calorimeter, the first tagged neutrino candidate.","lead":"NA62 has recorded what it calls the first tagged neutrino candidate: a neutrino whose parent kaon decay and energy are reconstructed from the charged particles. The single clean event is a proof of principle for a technique that could improve energy resolution in future neutrino experiments.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The background A estimate in Sec. 3 extrapolates a linear dLKr fit from the 60–300 mm sideband into the masked dLKr<60 mm signal region; if accidental activity is not uniform, the 0.034 background and the 3% background probability are not reliable.","rationale":"The paper's central claim is deliberately modest: one tagged-neutrino candidate, not a discovery. The observation is consistent with expected signal (0.208) plus background (0.034), and the analysis follows good practice: masked signal region, data-driven sidebands, explicit systematics, and honest uncertainties. My stress-test focused on the weakest link connecting the single event to the claim of a tagged neutrino: the background A model. The uniform-accidental-activity assumption is physically plausible, but it is exactly the kind of assumption that can hide a detector-related nonuniformity. Because the signal region is masked, the assumed linearity below 60 mm is never checked directly in data, and the sideband has limited statistics. I agree with the reader that this is the most fragile load-bearing premise. However, I do not see an internal inconsistency or a demonstrated error; the uncertainties are reported transparently, and the result is appropriately worded as a candidate. The proposed control-sample test would either validate the extrapolation or reveal a problem. Based on the current text, no verdict change is needed.","tokens_in":12575,"tokens_out":15445,"duration_ms":158241,"concrete_test":"Using the downscaled normalisation sample of K+→μ+ν decays (which records LKr information), compute for each event the distance dLKr between the predicted neutrino position and the nearest in-time LKr cluster over the full range 0–300 mm, without applying the interaction selection or the dLKr mask. Fit the resulting dLKr distribution down to 0 mm and test whether it is consistent with dN/ddLKr ∝ dLKr as extrapolated from the 60–300 mm sideband. If there is an excess or deficit below 60 mm, the uniformity assumption fails and the background A estimate must be revised. A complementary closure test: define a pseudo-signal window, e.g., 100<dLKr<160 mm, predict its yield from fits to 60–100 and 160–300 mm, and require agreement.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The single most load-bearing assumption is the model of accidental (background A) activity in Sec. 3. The authors assume accidental activity is uniform over the LKr front plane, so dN/ddLKr is proportional to dLKr, and fit a linear function to the dLKr sideband from 60 to 300 mm. The expected background of 0.030^{+0.041}_{-0.023} is obtained by integrating this line into the masked region dLKr<60 mm, scaled by the efficiency of the removed φ and dMUV3 requirements. This extrapolation is not directly validated: the signal region is masked, and the sideband contains only a handful of events, so the fit cannot distinguish the assumed ∝ dLKr shape from a line with a positive intercept or mild curvature near the origin. If the density of accidental clusters has any radial dependence (beam-related halo, local reconstruction inefficiencies, or LKr acceptance boundary effects), the linear relation can fail precisely in the dLKr<60 mm region. The quoted systematic uncertainty only compares fit ranges [60,300] mm and [60,180] mm; it does not vary the functional form or test the uniformity assumption. Since the central claim that the observed event is a tagged neutrino candidate depends on the expected background of 0.034 and the resulting 3% probability that the event is background, an unmodeled enhancement of accidental activity near signal-like positions would weaken the claim. This is a real soft spot, but not a demonstrated flaw: the modeling is standard and the claim is carefully hedged as a candidate.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The NA62 collaboration reports the observation of a single candidate event consistent with a K+ -> mu+ nu_mu decay followed by a charged-current nu_mu interaction in the liquid krypton calorimeter. The event is selected from 2022 data using a dedicated trigger line, and the signal region is masked until after the background evaluation. Backgrounds are estimated from data sidebands: 0.030 events from accidental activity (background A) and 0.0039 events from mis-reconstructed decays (background B), giving a total expected background of 0.034 events. The expected signal is 0.208 events. One event passes all criteria, with kinematic neutrino energy 52.09 GeV and missing mass consistent with zero. The paper concludes that this is the first tagged neutrino candidate and demonstrates the feasibility of the neutrino tagging technique.","tokens_in":12827,"tokens_out":7300,"duration_ms":67431,"significance":"If the candidate is genuine, this would be the first tagged neutrino ever observed and an important proof-of-principle for future tagged-neutrino beams. The analysis follows good practice: the signal region was masked before background evaluation, the background estimates are data-driven from sidebands, and the trigger and selection efficiencies are measured on control samples. The paper is appropriately cautious in using the word 'candidate' in most of the text, and it provides the key kinematic properties of the event. The result is significant for the long-term development of neutrino tagging, even though the statistical evidence is limited to a single event.","major_comments":[{"comment":"The assumed uniformity of accidental activity over the LKr front plane is the linchpin of the dominant background estimate. The dLKr distribution is fitted from 60 to 300 mm with a linear function, and the expected background of 0.030 events in the signal region (dLKr < 60 mm) is obtained by extrapolation. Because the signal region is masked and the sideband contains a limited number of events, the fit cannot distinguish the assumed dN/ddLKr proportional to dLKr shape from a line with a positive intercept or a mildly curved distribution. The quoted systematic uncertainty (0.004) only accounts for the choice of fit range, not the functional form or the uniformity assumption itself. Since the observed event has dLKr = 31.4 mm and the total background probability of 3% is dominated by this estimate, the authors should directly validate the uniformity assumption, for example by measuring the distance from the extrapolated neutrino position to the nearest LKr cluster in a control sample of K+ -> mu+ nu_mu decays that does not use the neutrino interaction selection, and by testing the stability of the background integral under alternative functional forms (e.g., adding a constant or quadratic term). Without such validation, the robustness of the 3% background probability is not established.","section":"Section 3, Background A"},{"comment":"The paper states that the neutrino energy reconstructed from kinematics has a relative uncertainty of 0.34%, but it does not describe how this value is derived from the GTK and STRAW momentum resolutions or from any momentum-scale systematic uncertainties. Since the sub-percent energy resolution is a central advertised benefit of the tagging technique, the source of this number should be specified, with a reference or a short derivation, so that the reader can assess whether the claim is justified.","section":"Section 5 and Table 2"}],"minor_comments":[{"comment":"The title uses 'First detection of a tagged neutrino' while the abstract and conclusions consistently use 'tagged neutrino candidate'; the wording should be aligned to avoid overstating the statistical significance.","section":"Title and Abstract"},{"comment":"The asymmetric statistical uncertainties on the background estimates are not fully explained; the paper should specify whether they arise from Poisson limits on the fitted yields, from the fit parameter uncertainties, or from the scale-factor uncertainty, and how the quoted central values are obtained.","section":"Section 3"},{"comment":"The sideband fits shown in Figure 3 contain only a small number of events; the paper should report the number of events used in each fit and the fit quality (e.g., chi2 per degree of freedom) so that the reader can judge the reliability of the extrapolation.","section":"Figure 3"},{"comment":"The statistical uncertainty on the expected signal Nsig = 0.208 +/- 0.013 is given as a single number, but the individual contributions from the normalisation sample, P_int, and the efficiency ratio are not broken out; a short breakdown would help identify the dominant source of uncertainty.","section":"Section 4"},{"comment":"The sentence 'Except for the signals associated with the neutrino candidate and the muon, no in-time LKr signal should be reconstructed' is slightly ambiguous; it would be clearer to state that no other LKr clusters in time with the event are allowed within the geometrical acceptance used for the analysis.","section":"Section 2, Interaction selection"}],"recommendation":"major_revision","confidential_remarks":"The manuscript reports a single candidate event with an expected background of 0.034 events, so the evidence is not a high-significance discovery. The main risk is the robustness of the dominant background A estimate, which relies on an extrapolation of a linear fit under an assumption of uniform accidental activity. I agree that the concern is a soft spot rather than a demonstrated flaw, but because it directly affects the 3% background probability that underlies the central claim, the authors should be required to provide additional validation or to temper the conclusion accordingly. The rest of the analysis appears careful and well documented."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First things first: this is a genuine first measurement, not a proposal or a simulation. NA62 has one event consistent with a K+ -> mu+ nu decay in the beam and a charged-current neutrino interaction in the LKr calorimeter, with the neutrino energy determined kinematically to 52.09 GeV at 0.34% precision. The analysis is careful: signal region masked, backgrounds from data sidebands, systematics evaluated. The claim is appropriately hedged as a 'candidate.'\n\nThe strongest part of the paper is the derivation of the expected signal, 0.208 events, from the normalization sample and GENIE. It is not fitted to the observed event, so there is no circularity. The background estimate for accidental activity (background A) is the softest spot. The paper assumes uniform accidental activity over the LKr front plane, giving dN/ddLKr proportional to dLKr, and fits a line to the 60–300 mm sideband. The signal region is dLKr < 60 mm, so the expected background of 0.034 events relies on an extrapolation of that linear model into a region where the fit has no direct leverage. The quoted systematic only compares fit ranges, not functional forms or the uniformity assumption. That concern is real, but it is not a demonstrated flaw: the sideband data are consistent with the model, and even a factor-of-two shift in this background would leave the event's interpretation plausible.\n\nThe total expected background is 0.034 events; the probability that the observed event is background is about 3 percent. That is suggestive, not definitive, and the paper does not overclaim. The result is a proof-of-principle for neutrino tagging, relevant to the community planning ENUBET, NuTAG, and similar efforts.\n\nI would send this to a serious referee. It is a solid experimental result with a clear, carefully worded claim; the main limitation is statistical, not methodological. A reader in neutrino physics will want to see it.","headline":"First tagged neutrino candidate: a careful single-event proof-of-principle whose main soft spot is the extrapolated accidental background.","tokens_in":13418,"tokens_out":2563,"would_cite":true,"duration_ms":23335,"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":"The NA62 experiment reports the first tagged neutrino candidate: a neutrino from a $K^+ \\to \\mu^+\\nu_\\mu$ decay is caught through a charged-current interaction in a liquid krypton calorimeter, its energy fixed kinematically to 52.09 GeV…","keywords":["neutrino tagging","tagged neutrino","charged-current interaction","liquid krypton calorimeter","kinematic energy reconstruction","kaon decay","NA62"],"falsifier":"Measure accidental activity below 60 mm using a control sample of $K^+\\to\\mu^+\\nu_\\mu$ decays selected without the neutrino-interaction requirement, counting unrelated liquid-krypton clusters as a function of distance from the predicted neutrino impact point. If the density below 60 mm is inconsistent with the linear extrapolation from the 60-300 mm sideband, the 0.034-event background is underestimated and the single candidate cannot be attributed to a tagged neutrino on the present evidence.","tokens_in":12289,"feed_emoji":"⚛️","tokens_out":12745,"duration_ms":108023,"temperature":0.7,"pith_summary":"The paper reports the first tagged neutrino candidate from the 2022 data sample of the NA62 experiment. The candidate is a $K^+ \\to \\mu^+\\nu_\\mu$ decay whose neutrino is detected through a charged-current interaction in the liquid krypton calorimeter, with the parent kaon and the muon fully reconstructed. If correct, this demonstrates that the neutrino tagging technique is experimentally feasible: a neutrino can be associated with its specific parent decay, and its energy can be obtained from decay kinematics rather than from the interaction final state. The neutrino energy is $E_\\nu = 52.09$ GeV with a relative uncertainty of 0.34%, against an expected signal of 0.208 events and an expected background of 0.034 events. The result is a proof-of-principle step toward accelerator-based neutrino experiments that use tagging to control flux, cross-section, and energy-scale systematics.","feed_headline":"One event shows neutrino tagging can work","feed_subtitle":"A neutrino from a kaon decay was caught in a krypton calorimeter, with its energy pinned to 0.34 percent.","key_machinery":"The central object is the tagged neutrino candidate: a neutrino event whose parent $K^+ \\to \\mu^+\\nu_\\mu$ decay is fully reconstructed, so the neutrino four-momentum is defined at the decay vertex as $P_{K^+} - P_{\\mu^+}$ under the kaon and muon mass hypotheses. The argument is carried by the combination of a high-resolution beam spectrometer for the parent kaon, a magnetic spectrometer for the muon, the liquid krypton calorimeter used as an active target for the charged-current interaction, and selection criteria that include a missing-mass cut $|m^2_{\\mathrm{miss}}| < 0.006$ GeV$^2/c^4$, a distance cut $d_{\\mathrm{LKr}} < 60$ mm between the predicted and reconstructed neutrino impact points, and coplanarity and muon-identification conditions. Kinematic reconstruction is the load-bearing piece: it gives a neutrino energy known to 0.34%, compared with the 10-30% typical of reconstruction from the measured interaction final state.","core_discovery":"On the paper's own terms, the central claim is that a single event observed in the 2022 NA62 data is a genuine $K^+ \\to \\mu^+\\nu_\\mu$ decay followed by a charged-current neutrino interaction inside the liquid krypton calorimeter. The neutrino direction and energy are not inferred from the interaction products; they come from the measured $K^+$ and $\\mu^+$ four-momenta, so the event is tagged. The squared missing mass is $m^2_{\\mathrm{miss}} = -0.00086$ GeV$^2/c^4$, consistent with zero, and the distance between the predicted neutrino impact point and the observed calorimeter signal is $d_{\\mathrm{LKr}} = 31.4$ mm. With 0.208 expected signal events and 0.034 expected background events, the probability of observing one event under the signal-plus-background hypothesis is 19%, while the probability under background-only is 3%. The paper concludes that the event is the first tagged neutrino candidate and that it demonstrates the feasibility of the tagging technique.","pith_inferences":["A direct continuation test is to re-run the identical masked selection on the 2023-2024 dataset: if the tagging interpretation is right, roughly 0.2 signal events per year should appear with the same $d_{\\mathrm{LKr}}$ and missing-mass signatures, and the sideband shape should remain linear; if the candidate is accidental background, the sidebands should grow proportionally.","The main scaling bottleneck for a dedicated tagged-neutrino facility is likely accidental activity rather than neutrino rate, so the uniform-activity assumption will need to be measured directly, for example with a dedicated control trigger, before the technique is used for precision physics.","Because the kinematic tag gives 0.34% energy resolution, tagged neutrino interactions could also serve as an in-situ energy-calibration source for large far detectors, independent of hadronic energy reconstruction.","The same tagging logic applies to pion decays in flight, which are more abundant than kaon decays and would extend the method beyond kaon parents."],"forward_implications":["The neutrino tagging technique is feasible with existing detectors: a neutrino from a specific kaon decay can be caught and identified in an instrumented calorimeter.","Tagged neutrino energies can be determined at sub-percent precision, far better than the 10-30% typical of final-state reconstruction.","A dedicated tagged-neutrino experiment could measure total and differential neutrino cross-sections with reduced flux and interaction-model systematics.","In long-baseline oscillation experiments, tagged neutrinos would improve energy resolution and reduce backgrounds from non-oscillated neutrinos.","The 2023-2024 NA62 data can be used to consolidate the single-candidate result."],"supporting_citations":[{"why":"Supplies the neutrino tagging method and the kinematic reconstruction whose feasibility this paper demonstrates.","marker":"[12]"},{"why":"Defines the beamline and detector layout used to reconstruct the kaon, the muon, and the calorimetric interaction.","marker":"[19]"},{"why":"Provides the detector simulation used to model the calorimeter response and evaluate selection efficiencies.","marker":"[21]"},{"why":"Supplies the neutrino event generator and its tunes used to compute the interaction probability and final states.","marker":"[22–25]"},{"why":"Gives the charged-current neutrino cross-section used to convert the decay sample into an expected signal.","marker":"[27]"},{"why":"Documents the trigger system and supplies the L1 efficiency entering the signal expectation.","marker":"[20]"}],"fun_headline_variants":["First tagged neutrino candidate seen in NA62","One kaon decay yields tagged neutrino signal","NA62 catches neutrino via kaon decay tag","Tagged neutrino detected for first time","Single event shows neutrino tagging works"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result rests on the background estimate, which assumes accidental calorimeter activity is uniform across the front plane so that a linear fit to the $d_{\\mathrm{LKr}}$ sideband from 60 to 300 mm can be extrapolated into the signal region below 60 mm; if accidental activity clusters near the predicted neutrino impact point, the true background could exceed the quoted 0.034 events.","fun_headline_variants_meta":{"raw":{"variants":["First tagged neutrino candidate seen in NA62","One kaon decay yields tagged neutrino signal","NA62 catches neutrino via kaon decay tag","Tagged neutrino detected for first time","Single event shows neutrino tagging works"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000139,"raw_usage":{"total_tokens":1093,"prompt_tokens":814,"completion_tokens":279,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":430,"completion_tokens_details":{"reasoning_tokens":215}},"tokens_in":430,"tokens_out":279,"duration_ms":2650,"temperature":1.0,"reasoning_tokens":215,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T21:49:54.871684+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure accidental activity below 60 mm using a control sample of $K^+\\to\\mu^+\\nu_\\mu$ decays selected without the neutrino-interaction requirement, counting unrelated liquid-krypton clusters as a function of distance from the predicted neutrino impact point. If the density below 60 mm is inconsistent with the linear extrapolation from the 60-300 mm sideband, the 0.034-event background is underestimated and the single candidate cannot be attributed to a tagged neutrino on the present evidence.","supporting_citations":[{"cited_title":"Allison, et al., Recent developments in Geant4, Nucl","cited_arxiv_id":null,"evidence_quote":"Provides the detector simulation used to model the calorimeter response and evaluate selection efficiencies."},{"cited_title":"Perrin-Terrin, Neutrino tagging: a new tool for accelerator based neutrino experiments, Eur","cited_arxiv_id":null,"evidence_quote":"Supplies the neutrino tagging method and the kinematic reconstruction whose feasibility this paper demonstrates."},{"cited_title":"Cortina Gil, et al., Performance of the NA62 trigger system, JHEP 03 (2023) 122","cited_arxiv_id":null,"evidence_quote":"Documents the trigger system and supplies the L1 efficiency entering the signal expectation."}],"review_version":1}