{"id":"f5a6db9a-0d66-4e5f-991e-1308817dc18f","arxiv_id":"2610.10185","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"At muTRISTAN's asymmetric mu+e- mode, projected 95% CL searches could exclude electron-only ALP and Z' couplings down to roughly 1e-5 at low mass and would be the only projected constraint above about 220 GeV.","lead":"The paper calculates how well the proposed muTRISTAN muon-electron collider could find new particles that interact only with electrons, using a four-lepton signal. It finds the collider could reach couplings as small as about ten millionths in a mass range that current experiments barely cover.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The quoted 10^-5 coupling reach is inconsistent with the paper's own Tables 2-3: using Eq. 4.3 and the tabulated B values gives g_aee ~ 6e-3 and g_Z' ~ 2e-3 at 5.5 GeV, not 6.5e-5 and 4.2e-5.","rationale":"In good faith, the paper presents a complete analysis chain: event generation, smearing, BDT pairing, Crystal Ball fits, and asymptotic significance. The tables are detailed enough to cross-check the central claim, and that cross-check fails badly. Table 2 at m_a = 5.5 GeV lists S = 3971 at g_aee = 0.1 with B = 60; the quoted 95% CL coupling 6.5e-5 would mean S ~ 0.0017 events, which cannot be excluded over 60 background events by any reasonable statistic. Solving the paper's own Eq. 4.3 gives S ~ 16 and hence g_aee ~ 6e-3. The Z' case is analogous: Table 3 gives S = 118 at g_Z' = 0.005 and B = 58, so the quoted 4.2e-5 requires S ~ 0.008 events; the actual solution is g_Z' ~ 1.8e-3. The inconsistency is largest at low mass, where the central 'order 10^-5' claim lives, but it also affects the high-mass endpoints: at m_a = 300 GeV the tabulated S_ref = 31 and B = 45 give a limit of order 6.6e-2 rather than 1.7e-2. Since the quoted cross-section limits and coupling limits are also mutually inconsistent with the tables, this is not a stylistic or interpretive issue but a quantitative error in the paper's main result. The reader's weakest assumption, the ILC-like detector and beam-induced backgrounds, is acknowledged by the authors and would affect the numbers at the factor-of-a-few level; the normalization inconsistency identified here is larger by two orders of magnitude and is not acknowledged. A corrected version would need to recompute all reach curves and would lose the claimed four-order-of-magnitude improvement at the light end, so the current version should not be accepted as a projection with the stated sensitivities.","tokens_in":27753,"tokens_out":28989,"duration_ms":202090,"concrete_test":"Recompute the 95% CL coupling limits from Tables 2 and 3 using Eq. 4.3: for each benchmark, solve Z_A(S,B) = 1.96 with S = S_ref * (g/g_ref)^2 and B taken from the table. At m = 5.5 GeV this gives g_aee ~ 6.3e-3 and g_Z' ~ 1.8e-3. If these recomputed values replace the quoted 6.5e-5 and 4.2e-5, the headline claim is refuted. As a cross-check, regenerate the m = 5.5 GeV benchmark with MadGraph at the quoted and the recomputed couplings and verify that the yielded signal counts reproduce the scaling assumed here.","verdict_should_be":"REJECT","load_bearing_attack":"The central sensitivity numbers do not follow from the paper's own event yields. Table 2 gives S = 3971 events at g_aee = 0.1 and B = 60 events at m_a = 5.5 GeV for L = 1000 fb^-1. With Eq. 4.3, excluding at 95% CL (Z_A = 1.96) requires about 16 signal events, so the reach should be g_aee = 0.1 * sqrt(16/3971) ~ 6.3e-3, roughly a factor of 100 weaker than the quoted 6.5e-5. For the Z' at 5.5 GeV, Table 3 gives S = 118 at g_Z' = 0.005 and B = 58; the required S ~ 15 gives g_Z' ~ 1.8e-3, about 40 times weaker than the quoted 4.2e-5. The same problem persists across the mass range: at m_a = 300 GeV, S_ref = 31 and B = 45 would give g_aee ~ 6.6e-2 instead of the stated 1.7e-2, and the stated cross-section limit of 0.03 fb at 5.5 GeV corresponds to S ~ 12 events, for which Eq. 4.3 gives Z_A ~ 1.5, below 1.96. Even allowing for the difference between limit-setting and discovery test statistics, no 95% CL exclusion with B = 60 can exclude an expected signal of 0.0017 events, which is what g_aee = 6.5e-5 would imply. The headline 'couplings of order 10^-5' and the claimed improvements over existing bounds are therefore not supported by the paper's own tabulated analysis.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper proposes using the asymmetric μ+e− mode of μTRISTAN at √s=346 GeV to search for electrophilic ALP and Z′ bosons in the e−e+e−μ+ final state, with the decay pair identified by a BDT and the e+e− invariant mass fitted with a double-sided Crystal Ball function. It reports projected 95% CL coupling limits down to about 6.5×10−5 for the ALP and 4.2×10−5 for the Z′, and claims improvements over existing bounds by several orders of magnitude, together with the only constraint above about 220 GeV. I checked the arithmetic connecting the quoted limits to the paper's own Tables 2 and 3 and Eqs. (4.2)–(4.3): the headline coupling limits do not follow from the tabulated signal and background yields. The tabulated yields imply low-mass coupling limits around 6×10−3 (ALP) and 1.8×10−3 (Z′), roughly one and a half to two orders of magnitude weaker than quoted, which would reverse the claimed comparison with the BaBar bounds.","tokens_in":28223,"tokens_out":10513,"duration_ms":69850,"significance":"The physics question is timely, and the paper is well organized: the common e−e+e−μ+ signature, the BDT-based pairing, the DSCB fitting, the peak-position checks, and the explicit estimate of signal–background interference are all sensible elements. The authors also honestly state the main external limitations, namely the parton-level simulation, the use of ILC/ILD-like detector resolution, and the neglect of beam-induced backgrounds from the stored muon beam. However, the central quantitative result, the order-10−5 coupling reach, is contradicted by the paper's own event yields, and the sections that compare with BaBar, LEP, and IceCube inherit that contradiction. As written, the significance of the paper is therefore not established, despite the soundness of much of the analysis pipeline around the incorrect limit-setting step.","major_comments":[{"comment":"The quoted ALP coupling reach does not follow from the paper's own yields. At m_a=5.5 GeV, Table 2 lists S=3971 signal events at g_aee=0.1 and B=60 background events for L=1000 fb−1. Solving Eq. (4.3) with Z_A=1.96 requires S≈16 events, giving g_aee≈0.1×sqrt(16/3971)≈6×10−3, not the quoted 6.5×10−5. The quoted coupling corresponds to about 0.0017 expected signal events over B=60, which cannot be excluded at 95% CL by any counting statistic. The same inconsistency appears at m_a=300 GeV, where S=31 and B=45 imply g_aee≈6×10−2 rather than the stated 1.7×10−2. The stated cross-section limit of 0.03 fb at 5.5 GeV corresponds to S≈12 events and gives Z_A≈1.5 from Eq. (4.3), below the 1.96 threshold.","section":"Section 4.1, Table 2, Eq. (4.3)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The same arithmetic failure occurs for the Z′. At m_Z′=5.5 GeV, Table 3 gives S=118 at g_Z′=0.005 and B=58. Requiring Z_A=1.96 needs S≈15–16 events, hence g_Z′≈0.005×sqrt(15/118)≈1.8×10−3, about 40 times larger than the quoted 4.2×10−5. At m_Z′=300 GeV, with S=0.19 and B=32, the implied coupling is about 4×10−2, not 1.1×10−2. These corrected values lie above, not below, the BaBar bound around 10−4 at low mass, so the claimed improvement over existing constraints is not supported by the tabulated yields.","section":"Section 4.2, Table 3, Eq. (4.3)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The interpretation section and the abstract inherit the inconsistency of the underlying limit calculation. The teal projected curves in Figs. 9 and 10 are compared with external exclusions, but those curves are generated from the same inconsistent numbers. Using the yields in Tables 2 and 3, the projected low-mass ALP limit is about 6×10−3, which is weaker than the BaBar bound by roughly two orders of magnitude, and the Z′ limit is similarly weaker; this inverts the paper's stated conclusion of an order-of-magnitude improvement. The claim of providing the only constraint above about 220 GeV may survive a corrected analysis, but the central quantitative claim of order-10−5 couplings does not.","section":"Section 5, Figs. 9–10; Abstract and Conclusion","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The text presents Eqs. (4.2) and (4.3) as the basis for 95% CL exclusion limits, but these are the Asimov discovery-significance formulas of refs. [102,103]. A 95% CL expected upper limit is normally obtained from the profile-likelihood q_μ test statistic or from a CL_s procedure. The authors should clarify why Z_A>1.96 defines a 95% CL exclusion; this matters for the precise meaning of 'sensitivity', although it is secondary to the numerical inconsistency above.","section":"Eqs. (4.2)–(4.3)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The contents lists both Section 1 and Section 2 as 'Introduction', and the body repeats the '2 Introduction' heading; this appears to be a compilation artifact and should be corrected.","section":"Table of contents and section headings","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The limitations concerning parton-level simulation, ILC-like detector resolution, and beam-induced backgrounds are stated clearly by the authors at the end of the paper. I do not treat these as new defects, but a revised version should quantify how they affect the corrected limits.","section":"Section 6, final paragraph","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"reject","confidential_remarks":"The central quantitative claim is contradicted by the paper's own tables and equations, so the manuscript should not be accepted in its current form. If the authors can trace the quoted limits to a corrected yield calculation, the paper could be reconsidered; otherwise the corrected couplings are roughly one and a half to two orders of magnitude weaker and the comparison with BaBar inverts."},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Check the quoted coupling limits against Tables 2 and 3 before believing the abstract. For the ALP at m_a = 5.5 GeV, Table 2 gives S = 3971 events at g_aee = 0.1 and B = 60 at L = 1000 fb^-1. Using the paper's own Eq. 4.3 with Z_A = 1.96, a 95% exclusion needs S ~ 16, so the coupling reach is 0.1 * sqrt(16/3971) ~ 6e-3, not the quoted 6.5e-5. The same arithmetic fails across the mass range: at m_a = 300 GeV, S_ref = 31 and B = 45 give g_aee ~ 6.6e-2 against the stated 1.7e-2. For the Z' at 5.5 GeV, Table 3 gives S = 118 at g = 0.005 and B = 58, which yields ~1.8e-3, not 4.2e-5. The quoted cross-section limit of 0.03 fb at 5.5 GeV implies S ~ 12 events, which with B = 60 gives Z_A ~ 1.5, below 1.96. So the headline 'couplings of order 10^-5' and the claimed improvements over BaBar, LEP, and IceCube are not supported by the paper's own numbers.\n\nThat is a serious problem, because the rest of the analysis is coherent and often careful. The novelty is application-level but real: no prior paper in the references quantifies electrophilic ALP and Z' sensitivity in muTRISTAN's asymmetric mu+e- mode. The BDT spectator tagging with invariant-mass variables excluded is sensible; the fitted peak positions reproduce generated masses to better than 0.15%; the Z->ee background bump at 90 GeV and the kinematic suppression at 300 GeV are handled consistently; and the authors state the EFT assumptions and the parton-level, ILC-smearing, and beam-induced-background limitations explicitly.\n\nMinor soft spots: 'providing the only constraint above ~220 GeV' is a projection, not a constraint, and should be reworded; the 'ATLAS' bound in Sec. 5 is actually ref. [25], a theory review, not an ATLAS paper. Both are easy fixes. The central inconsistency is not.\n\nWho this is for: people working on future-lepton-collider projections for leptophilic mediators. The EFT setup, final state, and analysis chain are worth preserving; the physics conclusions as written are not. I would send this to peer review rather than desk-reject, because the failure is localized and checkable and the corrected reach could still be of interest, but the authors need to redo the limit conversion and re-evaluate what remains of the claimed sensitivity. Once the numbers are corrected, I would reconsider citing it.","headline":"The quoted 10^-5 coupling reaches do not follow from the paper's own Tables 2 and 3; the rest of the analysis is coherent but this inconsistency is load-bearing.","tokens_in":29622,"tokens_out":1047,"would_cite":false,"duration_ms":54592,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"A proposed 346 GeV muon-electron collider mode could probe electron-coupled bosons far below current limits.","keywords":["electrophilic mediators","axion-like particles","Z-prime boson","muTRISTAN","asymmetric muon-electron collider","resonance search","four-lepton final state","projected exclusion limits"],"falsifier":"A detector-level simulation of the $\\mu^+e^-$ mode that includes beam-induced backgrounds (for instance from muon decays in the stored beam) and finds surviving extra events in the $e^-e^+e^-\\mu^+$ final state at the level of the ~22 fb Standard Model continuum would push the low-mass coupling reach above $10^{-4}$ and erase the claimed improvements; a simulation that reproduces the continuum within a few fb would confirm the projection.","tokens_in":29153,"feed_emoji":"⚛️","tokens_out":17164,"duration_ms":104590,"temperature":0.7,"pith_summary":"The paper aims to close the least-explored window for new bosons that couple only to electrons: mediator masses from roughly ten to a few hundred GeV, a region above the reach of the B factories that hadron colliders struggle with and that existing $e^+e^-$ data cover only partially. Its proposal is the asymmetric $\\mu^+e^-$ mode of the future $\\mu$TRISTAN collider, where a 1 TeV muon beam meets a 30 GeV electron beam at $\\sqrt{s} = 346$ GeV, putting an electron in the initial state from which the new boson can be radiated directly. The paper shows that both minimal realizations — a pseudoscalar axion-like particle and a $Z'$ vector boson — give the same clean four-lepton signature, $e^-\\mu^+ \\to e^-e^+e^-\\mu^+$, and that a resonance search in the $e^+e^-$ invariant mass for masses between 5.5 and 300 GeV reaches 95% CL couplings of order $10^{-5}$ at the light end. If these projections hold, the region between the kinematic limits of the B factories and the LEP contact-interaction bounds, currently the least constrained part of the electrophilic-mediator parameter space, would become testable at a lepton collider with a hadron-free initial state.","feed_headline":"Muon-electron collisions could unmask electron-only bosons","feed_subtitle":"At 346 GeV, the asymmetric μ+e− mode beats today's bounds by up to four orders of magnitude.","key_machinery":"The load-bearing object is the asymmetric beam configuration itself: the $\\mu^+e^-$ mode of $\\mu$TRISTAN with $E_e = 30$ GeV against $E_\\mu = 1$ TeV gives $\\sqrt{s} = 346$ GeV with an electron in the initial state, so an electrophilic mediator can be produced at tree level in a hadron-free environment. The common signature is the four-lepton final state $e^-\\mu^+ \\to e^-e^+e^-\\mu^+$; because both the ALP and the $Z'$ decay promptly to $e^+e^-$, one analysis chain serves both spins. That chain rests on three pieces: a boosted decision tree (the XGBoost implementation) that tags the spectator electron using nineteen production-level kinematic variables while excluding $m_{ee}$ itself, so the tagging cannot sculpt the discriminant; a double-sided Crystal Ball fit to the $m_{ee}$ peak that defines the signal window $[\\mu-\\sigma, \\mu+\\sigma]$ for each benchmark; and the Asimov significance formula that converts signal and background yields into 95% CL limits, with detector response emulated by Gaussian momentum smearing tuned to the ILD concept.","core_discovery":"Both mediators analyzed here are produced by radiation off the incoming electron and decay promptly back into an electron-positron pair, so the two scenarios share one narrow-resonance signature: a peak in the reconstructed $e^+e^-$ invariant mass on top of a smooth photon/$Z$-mediated continuum. The paper's central technical ingredient is identifying which of the two final-state electrons is the decay partner of the positron; a boosted-decision-tree tagger, trained on production kinematics alone with $m_{ee}$ deliberately excluded, selects the correct pair in 83% of events on average for the ALP and 75% for the $Z'$ across the full mass range, where no single kinematic rule works. Fitting the resulting peak with a double-sided Crystal Ball function and applying an Asimov significance formula, the paper derives 95% CL exclusion limits for fifteen benchmark masses at four luminosities and two systematic assumptions. With 1000 fb$^{-1}$, the coupling reach is $g_{aee} \\lesssim 6.5\\times10^{-5}$ and $g_{Z'} \\lesssim 4.2\\times10^{-5}$ at the light end, degrading to the $10^{-2}$ range at 300 GeV; the degradation is set by the phase-space fall-off of the production cross-section, not by analysis performance. Against existing bounds, the projections are comparable to the BaBar search below about 10 GeV, surpass all current constraints above it — by more than four orders of magnitude for the ALP relative to perturbative-unitarity limits — and provide the only constraint on the heaviest $Z'$ masses above roughly 220 GeV.","pith_inferences":["The $\\mu^+\\mu^+$ mode at $\\sqrt{s} = 2$ TeV, which the paper explicitly leaves to future work, could carry the same four-lepton search above 300 GeV once photon-initiated electron pairs supply the radiating electron, at the price of extra powers of the electromagnetic coupling.","The honest reading of the quoted reach inherits the paper's own caveats: the detector response is an ILC-like stand-in and muon-beam-induced backgrounds are not simulated, so a dedicated detector study could move the limits in either direction, and how far they move decides whether the four-order-of-magnitude improvement over current bounds is realised in practice.","The recipe here — an energetic electron beam in an asymmetric lepton collider, a resonance-blind tagger, and a narrow invariant-mass window — generalizes to any future electron-beam machine, meaning the search strategy, not just this machine, is the transferable asset."],"forward_implications":["At 1000 fb$^{-1}$ and null systematics, an electrophilic ALP with $5.5 \\le m_a \\le 100$ GeV and $g_{aee}$ between $6.5\\times10^{-5}$ and $2.7\\times10^{-3}$ would be excluded at 95% CL, a region that no current experiment constrains directly.","The $Z'$ reach, $g_{Z'} \\lesssim 4.2\\times10^{-5}$ at low mass, would improve on the LEP mono-photon, LEP-2 contact-interaction, and IceCube bounds by up to two orders of magnitude, and above $m_{Z'} \\simeq 220$ GeV the $\\mu^+e^-$ search would be the only constraint on the parameter space.","Since both scenarios share the same final state, reconstruction chain, and statistical treatment, a single run simultaneously probes pseudoscalar and vector electron couplings across 5.5–300 GeV.","A 5% systematic uncertainty on the background degrades the coupling limits by at most about 25%, so the headline reach does not hinge on precise background control as long as the 22.3 fb continuum estimate is right.","The reach is dominated by the production cross-section rather than the selection, so the same pipeline extends to other beam energies and mediator masses without redesign."],"supporting_citations":[{"why":"Supplies the machine design and parameters (1 TeV muon beam against 30 GeV electrons, $\\sqrt{s}=346$ GeV, up to 1000 fb$^{-1}$) on which the entire projection rests.","marker":"[82]"},{"why":"Precursor study showing an energetic electron beam can probe electrophilic bosons in the intermediate-mass window; it motivates the $\\mu$TRISTAN analysis and supplies the comparison point for the approach.","marker":"[81]"},{"why":"The BaBar $e^+e^-$ search whose low-mass exclusion the projections match below about 10 GeV and far exceed above it.","marker":"[53]"},{"why":"LEP/DELPHI mono-photon and contact-interaction bounds that constitute the main existing $Z'$ constraints the projection improves upon by up to two orders of magnitude.","marker":"[113, 114]"},{"why":"IceCube bound on non-standard neutrino interactions that constrains the neutrino coupling of anomaly-free $Z'$ realisations and competes with the projection at low mass.","marker":"[115]"},{"why":"ILC/ILD detector documents that supply the Gaussian momentum-resolution parameters used to emulate the $\\mu$TRISTAN detector response.","marker":"[97, 98]"},{"why":"Asimov significance formula that converts signal and background yields into the quoted 95% CL cross-section and coupling limits.","marker":"[102, 103]"},{"why":"Perturbative-unitarity bound on the ALP-electron coupling that the projected ALP limit beats by more than four orders of magnitude at low mass.","marker":"[112]"}],"fun_headline_variants":["Muon-electron collisions at 346 GeV could reveal electron-only bosons","Electrophilic boson searches at muTRISTAN's asymmetric μ+e− mode","ALP and Z' mediators via four-lepton resonance at muTRISTAN","At 346 GeV, μ+e− collisions could improve ALP coupling limits 10,000-fold","Search for electron-coupled bosons in e+e− pairs at muTRISTAN"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The projections assume that a detector for $\\mu$TRISTAN, which does not exist yet, will measure lepton momenta with the resolution of the ILC's ILD concept and that backgrounds from the stored muon beam will not add events to the $e^-e^+e^-\\mu^+$ final state; the authors acknowledge that only a dedicated detector design and a beam-background study can settle this.","fun_headline_variants_meta":{"raw":{"variants":["Muon-electron collisions at 346 GeV could reveal electron-only bosons","Electrophilic boson searches at muTRISTAN's asymmetric μ+e− mode","ALP and Z' mediators via four-lepton resonance at muTRISTAN","At 346 GeV, μ+e− collisions could improve ALP coupling limits 10,000-fold","Search for electron-coupled bosons in e+e− pairs at muTRISTAN"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000682,"raw_usage":{"cost":0.000682136,"is_byok":false,"cost_details":{"upstream_inference_cost":0.000682136,"upstream_inference_prompt_cost":0.000110656,"upstream_inference_completions_cost":0.00057148},"total_tokens":3241,"prompt_tokens":1200,"completion_tokens":2041,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":512,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":1928}},"tokens_in":1200,"tokens_out":2041,"cost_source":"openrouter_usage","duration_ms":13593,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":1928,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-08T15:21:17.608118+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"falsifier":"A detector-level simulation of the $\\mu^+e^-$ mode that includes beam-induced backgrounds (for instance from muon decays in the stored beam) and finds surviving extra events in the $e^-e^+e^-\\mu^+$ final state at the level of the ~22 fb Standard Model continuum would push the low-mass coupling reach above $10^{-4}$ and erase the claimed improvements; a simulation that reproduces the continuum within a few fb would confirm the projection.","supporting_citations":[{"cited_title":"$\\mu$TRISTAN","cited_arxiv_id":"2201.06664","evidence_quote":"Supplies the machine design and parameters (1 TeV muon beam against 30 GeV electrons, $\\sqrt{s}=346$ GeV, up to 1000 fb$^{-1}$) on which the entire projection rests."},{"cited_title":"ALP and $Z^\\prime$ boson at the Electron-Ion collider","cited_arxiv_id":"2601.04962","evidence_quote":"Precursor study showing an energetic electron beam can probe electrophilic bosons in the intermediate-mass window; it motivates the $\\mu$TRISTAN analysis and supplies the comparison point for the approach."},{"cited_title":"Hunting for ALPs with Lepton Flavor Violation","cited_arxiv_id":"1911.06279","evidence_quote":"Perturbative-unitarity bound on the ALP-electron coupling that the projected ALP limit beats by more than four orders of magnitude at low mass."}],"review_version":1}