{"id":"1396b7a7-75cb-4127-ae83-7c324944c73a","arxiv_id":"2502.04241","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"No hadronic decays of feebly-interacting particles are observed in NA62 beam-dump data, excluding new regions of dark photon, dark scalar, and axion-like particle parameter space.","lead":"The NA62 experiment searched for dark photons, dark scalars, and axion-like particles decaying to hadrons in 1.4x10^17 proton-on-dump collisions and found no signal. The null result excludes new regions of dark sector parameter space beyond earlier experiments.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exclusion contours that rely on the approximate meson-mixing kinematics in Sec. 3.3 may be overconfident; the paper explicitly leaves this production channel without a complete treatment.","rationale":"The paper is a solid experimental search: zero observed events, small validated backgrounds, masked SR/CR, and standard CLs limits. The central null result is not in question. The only defensible soft spot is the model-dependent conversion from observed counts to exclusion contours, specifically the meson-mixing kinematics admitted to be incomplete. The reader identified the broader unpropagated theory uncertainty; I sharpen to the particular admitted approximation and propose a computation that would decide whether it matters. Because the paper is transparent and the zero-event statement stands, my stress-test does not alter the accept verdict; it only points to a qualification that should be kept in mind when citing the contours.","tokens_in":14212,"tokens_out":15170,"duration_ms":189784,"concrete_test":"Recompute the BC1 and BC11 observed 90% CL contours using the public ALPINIST framework with the meson-mixing production process disabled (or with NP momentum distribution replaced by the unmodified parent-meson momentum) and compare with the right panels of Figs. 7 and 9. If the contours outside the previous experimental limits are unchanged, the mixing approximation is not load-bearing; if they move inward by more than the 20% N_POT band, the 'new regions' claim should be explicitly qualified as dependent on an unvalidated kinematic approximation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim has two parts: the null observation (robust) and the new exclusion regions (model-dependent). The least secure ingredient in the latter is the meson-mixing production channel used for dark photons and ALPs. Section 3.3 states that 'there is no complete treatment of this effect available in the literature. Therefore, the NP particle kinematics is approximately evaluated as discussed in Ref. [16].' The signal acceptance A_acc and decay-in-FV probability P_RD in Eq. (1) are functions of the NP-particle momentum and direction; the SR is defined in (Z_TAX, CDA_TAX), so an approximate kinematic redistribution changes which simulated events enter the SR and thus changes N_exp. The right panels of Figs. 7 and 10 include this mixing channel; if the approximation overestimates N_exp, the 90% CL contours stretch to smaller couplings, enlarging the claimed new regions. The paper's systematic budget covers only N_POT (20%) and A_acc statistics/PID (3.3%); no uncertainty is assigned to the mixing model itself. This is an explicit admitted limitation, not a hidden flaw, but it is exactly where the central 'new regions' claim is least supported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a search for long-lived feebly-interacting particles (dark photons, dark scalars, and axion-like particles) decaying to hadronic final states in the NA62 beam-dump configuration. Using a 2021 dataset corresponding to (1.4±0.3)×10^17 protons on the TAX dump, the analysis selects events with two oppositely charged hadronic tracks and possible photons, reconstructing final states such as π+π−, π+π−γ, π+π−π0, π+π−π0π0, π+π−η, K+K−, and K+K−π0. The signal and control regions are kept masked until background estimates are validated. The expected background in the signal region is below 0.04 events for all channels, dominated by K+ → π+π+π− decays; zero events are observed after unmasking. The null result is interpreted with the public ALPINIST framework to set 90% CL exclusion limits for several benchmark models, combining the hadronic channels with previously published di-lepton results. The paper reports new excluded regions for dark scalars, dark photons, and axion-like particles.","tokens_in":14443,"tokens_out":7470,"duration_ms":85981,"significance":"The central experimental observation—zero candidates against a validated background expectation below 0.04 events—is robust and constitutes the paper's main result. The analysis is carefully blinded, the background is estimated with data-driven methods (notably via single-track Q1 samples and PUMAS+GEANT4 simulation), and the systematic budget is clearly presented. The exclusion limits improve on previous beam-dump searches in significant parts of the parameter space, and the use of the public ALPINIST framework makes the interpretation reproducible. The main limitation is that the limits inherit the theoretical uncertainties of the production models, and in particular the meson-mixing channel for dark photons and ALPs is treated with an approximation for which no complete literature treatment exists; this is explicitly acknowledged in Sec. 3.3. This limitation does not affect the null observation and is normal for model-dependent searches.","major_comments":[],"minor_comments":[{"comment":"The meson-mixing production channel for dark photons and ALPs uses an approximate treatment with no assigned systematic uncertainty; since the contours in the right panels of Figs. 7 and 10 include this channel, the authors should add a sentence in Section 5 or in the figure captions stating explicitly that those exclusion contours are subject to this approximation.","section":"Section 3.3, Eq. (1)"},{"comment":"The mass axis is labelled 'mA′ [MeV/c2]' but the plotted range and context indicate that the label should be GeV/c2 rather than MeV/c2.","section":"Figure 6, bottom panel"},{"comment":"The axis labels are rendered inconsistently (e.g., ' [m] TAX Z' and ' [mm] TAX CDA'); please use the same notation as in the text, namely Z_TAX and CDA_TAX.","section":"Figure 3"},{"comment":"The statement that 'new regions ... are excluded' would benefit from a brief reminder that the exclusions are model-dependent and do not include theory uncertainties in the production processes, as already acknowledged in Section 3.3 and in the figure captions.","section":"Section 5 and Abstract"}],"recommendation":"minor_revision","confidential_remarks":"For the editor: the paper is a solid experimental search with a transparent analysis. The only substantive caveat is the approximate meson-mixing production model, which is clearly flagged in the text; I would ask the authors to add a one-sentence reminder in the conclusions. I do not see any circularity or authorship issues that would affect the decision, though the ALPINIST framework is developed in part by collaboration members; it is a published, public tool."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is a clean, well-executed null result from NA62's beam-dump run, and it is the first NA62 search for hadronic decays of feebly-interacting particles. The central observation — zero events in the signal region with an expected background below 0.04 — is robust and data-driven. The paper's real value is the combination of this hadronic search with the earlier di-lepton search, which extends the excluded parameter space for dark photons, dark scalars, and ALPs beyond previous beam-dump and collider limits in several benchmark models.\n\nThe analysis is careful. Backgrounds are estimated with a mix of data-driven methods and simulation: kaon decays, prompt muon interactions, combinatorial, and neutrino-induced are all studied, and the dominant K+ -> pi+ pi+ pi- background is validated by inverting vetoes. The signal region is masked until the background model is frozen. That's the right way to do this.\n\nThe soft spot is exactly where the stress test points: the meson-mixing production channel for dark photons and ALPs. Section 3.3 says plainly that there is no complete treatment in the literature and the kinematics are approximated as in Ref. [16]. That approximation feeds into the acceptance and decay-in-fiducial-volume probability, so the expected number of signal events — and therefore the exclusion contour — inherits an unquantified theory uncertainty. The paper does not propagate any theory uncertainty into the limits; the error bands only cover the 20% POT uncertainty and 3.3% acceptance systematic. For the mixing channel, the contours in the right panels of Figs. 7 and 10 may be overconfident if the approximation overestimates N_exp. I'd call that a caveat, not a flaw: the paper states it explicitly, and it also shows the limits without the mixing production (left panels), so the reader can see how much of the new exclusion region actually depends on the approximation. The null result itself does not depend on any of this.\n\nThe citation pattern looks healthy; the ALPINIST framework and earlier NA62 papers are cited, and the theory references are standard. The paper is an incremental but genuinely new experimental result, and it is honest about its own limitations.\n\nBottom line: this is a solid paper for the dark-sector community. The null result stands, the exclusion limits are useful with the stated caveat, and the analysis is done properly. I'd send it to peer review and accept it, possibly with a minor request to add one sentence noting that the meson-mixing contours carry an additional unquantified model uncertainty. It deserves a serious referee.","headline":"Clean first NA62 hadronic beam-dump search; the null result is robust, and the exclusion contours carry an admitted but unquantified theory uncertainty in the meson-mixing channel.","tokens_in":14926,"tokens_out":2303,"would_cite":true,"duration_ms":26776,"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":"Recording $1.4\\times10^{17}$ protons on dump, NA62 observes zero events in seven hadronic final states and excludes new dark photon, dark scalar, and axion-like particle parameter space.","keywords":["NA62","beam dump","dark photon","dark scalar","axion-like particle","feebly-interacting particles","hadronic decays","exclusion limits"],"falsifier":"Measure the inclusive B-meson production cross section in 400 GeV proton–nucleus collisions and carry out a complete meson-mixing calculation for dark-photon and axion production; if either lowers the expected signal counts by more than the 20% proton-on-target uncertainty, the quoted exclusion contours would move inward, and a future higher-statistics beam-dump run would be needed to restore them.","tokens_in":14035,"feed_emoji":"⚛️","tokens_out":9646,"duration_ms":94013,"temperature":0.7,"pith_summary":"NA62 reports a search for feebly-interacting particles produced by dumping 400 GeV protons onto an absorber, using $1.4\\times10^{17}$ protons on dump collected in 2021. Seven hadronic final states are studied: $\\pi^+\\pi^-$, $\\pi^+\\pi^-\\gamma$, $\\pi^+\\pi^-\\pi^0$, $\\pi^+\\pi^-\\pi^0\\pi^0$, $\\pi^+\\pi^-\\eta$, $K^+K^-$, and $K^+K^-\\pi^0$. After unmasking the signal and control regions, zero events are observed, against expected backgrounds below $0.05$ events in every channel. Combining this null result with the earlier di-lepton search on the same data excludes new regions of dark photon, dark scalar, and axion-like particle parameter space, extending beyond previous beam-dump and collider limits.","feed_headline":"NA62 beam-dump search sees zero hadronic dark-sector decays","feed_subtitle":"Seven final states, from ππ to KKπ0, close new windows on dark photons, dark scalars, and axions.","key_machinery":"The signal yield is governed by $N_{\\rm exp} = N_{\\rm POT}\\,\\chi^i_{pp\\to X}(m_X)\\,P^i_{\\rm RD}(m_X,\\tau_X)\\,A^{ij}_{\\rm acc}(m_X,\\tau_X)$, where $\\chi^i_{pp\\to X}$ is the production probability per proton on dump for production process $i$ (B-meson decays, light-meson decays and mixing, bremsstrahlung, or Primakoff), $P^i_{\\rm RD}$ is the probability that the particle traverses the dump and decays inside the 105–180 m fiducial volume, and $A^{ij}_{\\rm acc}$ is the acceptance for decay channel $j$. The hadronic branching ratios are set to unity at the reference stage, making the measured event counts model-independent before interpretation; the model-dependent exclusion contours are then produced with a public signal-yield and likelihood tool [16] using the CLs method [23]. Signal and control regions in the $(Z_{\\rm TAX},\\mathrm{CDA}_{\\rm TAX})$ plane, derived from the backward-extrapolated production vertex, are masked until the background estimates are validated.","core_discovery":"The central claim is that the NA62 beam-dump data contain no evidence of new particles decaying hadronically, and that this absence excludes parameter space that previous searches had left open. In the dark photon benchmark, the 90% CL observed contour in the $(m_{A'},\\varepsilon)$ plane combines hadronic and di-lepton channels and reaches couplings below those probed by earlier proton and electron beam-dump experiments for sub-GeV masses. The dark scalar limit on $\\sin^2\\theta$ is similarly extended, and for axion-like particles new limits are set on fermion and gluon couplings assuming a UV scale $\\Lambda=1$ TeV. In all benchmark cases the observed contours extend beyond the previous limits, with mass windows around the $\\pi^0$, $\\eta$, $\\eta'$, $\\rho$, $\\omega$, and $\\phi$ resonances excluded from the interpretation because the expected signal is not reliable there.","pith_inferences":["If the assumed production rates are accurate, the null result disfavors dark photons and dark scalars as the dominant dark-sector mediators over the newly excluded mass–coupling region, narrowing the parameter space for models that use these portals.","Because the analysis is essentially background-free at $10^{18}$ protons on dump, a larger dataset would gain mainly from statistics; production-model uncertainties, which are not propagated into the contours, would then become the limiting systematic.","A complete calculation of light-meson mixing kinematics for dark photons and axion-like particles could shift the contours, so the quoted excluded regions may need revision once such a treatment exists.","The seven hadronic final states are defined model-independently, so the same event selection can be reused to constrain any new particle decaying to those final states, not only the three benchmark models considered here."],"forward_implications":["Dark-photon and dark-scalar 90% CL contours now cover coupling values below the reach of earlier proton and electron beam-dump experiments for sub-GeV masses.","The gluon-coupled axion-like-particle benchmark is constrained almost entirely through hadronic channels, since its di-lepton widths are small.","The dominant background, $K^+\\to\\pi^+\\pi^+\\pi^-$, contributes only about 0.007 expected events in the signal region, so the search is not background-limited at the current sample size.","Updating the light-meson production spectra and adding mixing and time-like form-factor production also strengthens the earlier NA62 di-lepton-only bounds."],"supporting_citations":[{"why":"Defines the NA62 beam-dump configuration and reports the earlier di-lepton search whose dataset and production assumptions this analysis builds on.","marker":"[3]"},{"why":"Provides the di-lepton search whose results are combined here and whose profile-likelihood test statistic and signal-region definitions are reused.","marker":"[4]"},{"why":"Supplies the B-meson production cross section used to compute dark-scalar and axion production via B→KX decays.","marker":"[12]"},{"why":"Provides the decay widths for the axion-like-particle model interpretation.","marker":"[13]"},{"why":"Provides the decay widths and B→KX formulation for the dark scalar interpretation.","marker":"[14]"},{"why":"Validates the PYTHIA8 light-meson spectra used for dark-photon production.","marker":"[15]"},{"why":"Public software that combines production and decay channels and supplies the approximate meson-mixing kinematics used when no complete treatment exists.","marker":"[16]"},{"why":"Supplies the quasi-real approximation and form factors for dark-scalar bremsstrahlung production.","marker":"[17]"},{"why":"Supplies the Primakoff differential cross section used for axion-like-particle production.","marker":"[18]"},{"why":"Defines the CLs method used to derive the 90% confidence-level exclusion contours.","marker":"[23]"}],"fun_headline_variants":["NA62 beam dump excludes new dark particle decays","Zero hadronic dark decays in NA62 beam-dump data","NA62 tightens limits on dark photons and scalars","No sign of dark hadrons in NA62 proton dump","Hadronic dark-sector search at NA62 yields null result"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The limits stand on how often the dump is assumed to produce dark photons, dark scalars, and axion-like particles; if the true production rates are lower than assumed, the excluded regions shrink, and the paper does not fold theory uncertainty into the contours.","fun_headline_variants_meta":{"raw":{"variants":["NA62 beam dump excludes new dark particle decays","Zero hadronic dark decays in NA62 beam-dump data","NA62 tightens limits on dark photons and scalars","No sign of dark hadrons in NA62 proton dump","Hadronic dark-sector search at NA62 yields null result"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000727,"raw_usage":{"total_tokens":3208,"prompt_tokens":849,"completion_tokens":2359,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":465,"completion_tokens_details":{"reasoning_tokens":2279}},"tokens_in":465,"tokens_out":2359,"duration_ms":16867,"temperature":1.0,"reasoning_tokens":2279,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T23:01:18.860759+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the inclusive B-meson production cross section in 400 GeV proton–nucleus collisions and carry out a complete meson-mixing calculation for dark-photon and axion production; if either lowers the expected signal counts by more than the 20% proton-on-target uncertainty, the quoted exclusion contours would move inward, and a future higher-statistics beam-dump run would be needed to restore them.","supporting_citations":[{"cited_title":"Cortina Gil, et al., JHEP09, 035 (2023)","cited_arxiv_id":null,"evidence_quote":"Defines the NA62 beam-dump configuration and reports the earlier di-lepton search whose dataset and production assumptions this analysis builds on."},{"cited_title":"Schubert, B","cited_arxiv_id":null,"evidence_quote":"Supplies the B-meson production cross section used to compute dark-scalar and axion production via B→KX decays."},{"cited_title":"Bauer, M","cited_arxiv_id":null,"evidence_quote":"Provides the decay widths for the axion-like-particle model interpretation."},{"cited_title":"Jerhot, et al., JHEP07, 094 (2022)","cited_arxiv_id":null,"evidence_quote":"Public software that combines production and decay channels and supplies the approximate meson-mixing kinematics used when no complete treatment exists."}],"review_version":1}