{"id":"b584bd30-7009-4b60-83b2-ed50097e3e58","arxiv_id":"2511.00687","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Beam-induced electromagnetic deflection would shift CEPC's Z-pole luminosity count by ~6×10⁻³ if uncorrected — about 60 times the 10⁻⁴ precision target — with corrections computable from simulation and anchored on a µrad-level crossing-angle measurement.","lead":"Simulations of collisions at the proposed CEPC electron-positron collider show that the electromagnetic fields of the beams deflect both the incoming particles and the Bhabha-scattered electrons they are meant to count, distorting the luminosity measurement by about 6 parts per 1000 if uncorrected. That is roughly 60 times the design precision goal, so the work maps an obstacle the CEPC physics program must correct for.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The simulation-derived EMD1/EMD2 correction magnitude lacks an independent calibration; an unvalidated ~2× error in the GuineaPig kick would exceed the 10^-4 precision goal, and the ±10% parameter scans do not probe this.","rationale":"The reader's weakest assumption identified the GuineaPig kick and the boost/rotation association as the key unvalidated link; I agree this is load-bearing. My concern is slightly more specific: the paper's stability tests (Figs. 7, 11) vary beam parameters but cannot bound model error in the kick normalization, and EMD1 lacks the slice-convergence check that EMD2 receives. The unit typo (140 mrad vs 140 µrad) is real but cosmetic. The direction and order of magnitude of the claimed effects are plausible and consistent internally (e.g., 52 keV from E·tan(α/2)·Δα), so the verdict should remain CONDITIONAL, not REJECT. No change to the reader's verdict is needed.","tokens_in":6000,"tokens_out":3671,"duration_ms":41774,"concrete_test":"Independently compute the EMD1 transverse kick for the CEPC nominal parameters by integrating the Lorentz-transformed electric and magnetic fields of a Gaussian bunch along straight trajectories, obtaining the mean p_x kick per beam, and compare to GuineaPig's ~2.9 MeV. If the analytic value differs from 2.9 MeV by more than 10%, the EMD1 correction and its quoted 5×10^-4 systematic uncertainty are not validated; the correction should be re-derived with a code cross-check or an alternative field-solver setting.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central correction factors—EMD1 ΔLint/Lint ≈ 4×10^-3 and EMD2 ≈ 1.4×10^-3—are extracted entirely from GuineaPig V1.2.2 simulations, with the EMD1 chain also relying on a per-event boost/rotation of BHLUMI final states by the simulated initial-state kick. The quoted systematic uncertainties (5×10^-4 and 3×10^-4) are obtained by varying nominal beam parameters within ±10% (Figs. 7 and 11), but this exercises only the sensitivity around the simulated point, not the absolute accuracy of the kick strength or of the boost/rotation prescription. If the true mean kick were, say, half or twice the simulated 5.8 MeV, the correction error would be ~2–4×10^-3, orders of magnitude above the 10^-4 goal. The paper itself flags simulation-setting sensitivity for EMD2 (Section 4, Fig. 9), yet no equivalent convergence test is shown for EMD1, and no independent analytic estimate or second beam-beam code is used to validate the absolute scale. The 'loss is stable under ±10% beam-parameter variations' statement is therefore not evidence that the correction itself is correct; it only bounds variation around the unvalidated central value.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper estimates two electromagnetic deflection effects on the CEPC Z-pole low-angle Bhabha luminometer: EMD1 (deflection of initial-state electrons/positrons by the opposing bunch fields) and EMD2 (deflection of final-state Bhabha particles). Using GuineaPig V.1.2.2 and BHLUMI V4.04 with post-CDR CEPC beam parameters, it claims that if uncorrected the combined relative loss of Bhabha count is about 6×10^-3 (about 4×10^-3 from EMD1 and 1.4×10^-3 from EMD2), that the corrections can be determined from simulation with residual systematic uncertainties no larger than 5×10^-4 (EMD1) and 3×10^-4 (EMD2) under ±10% beam-parameter variations, and that a ~1 µrad measurement of the crossing angle from di-muon data could anchor the EMD1 correction. Section 2 states the assumptions of no beam-energy spread and no radiative processes; Section 5 discusses correction concepts.","tokens_in":6182,"tokens_out":5932,"duration_ms":66608,"significance":"If the numerical results are correct, the paper addresses a central systematic issue for the CEPC luminosity program and would be a useful first estimate. The strengths are that the forward simulation chain is clearly described, an internal-consistency check (∆E ≈ E·tan(α/2)·∆α giving ~52 keV) is presented, the EMD2 result is checked against longitudinal-slice-count convergence (Fig. 9), and the sensitivity to ±10% beam-parameter variations is explored. The main risk is that the absolute scale of both corrections comes from a single beam-beam simulation code with no independent calibration or analytic cross-check; the paper itself flags this sensitivity through reference [3]. Because the quoted systematic uncertainties are close to the 10^-4 goal, the missing external validation is load-bearing.","major_comments":[{"comment":"The EMD1 correction (∼4×10^-3) is derived by taking the GuineaPig initial-system kick (mean ∼5.8 MeV, crossing-angle reduction ∼140 µrad) and applying a per-event boost/rotation to BHLUMI final states. No independent analytic estimate or second beam-beam code is provided. The ±10% beam-parameter scan of Fig. 7 probes only the sensitivity around the simulated point, not the absolute accuracy of the kick or of the 'association' mapping. If the kick strength or the boost/rotation prescription is wrong by a factor of two, the correction error would be several times 10^-3, far above the 10^-4 goal. Please add an independent estimate of the mean kick (e.g., an analytic bunch-field calculation) and validate the association prescription by comparing with Bhabha events generated directly in the kicked initial-state frame.","section":"§3, Figs. 2–7"},{"comment":"The text says the bunch population was updated from 8×10^10 to 14×10^10 particles per bunch, but Table 1 lists N = 15×10^10. Since all quoted correction values scale with N and the ±10% variations are taken around this nominal value, this inconsistency must be resolved. Please state unambiguously which post-CDR parameter set was actually used and confirm that the quoted corrections correspond to that set.","section":"Introduction and Table 1"},{"comment":"The EMD2 simulation is performed without radiative processes and without beam-energy spread, and the ISR loss is imported from the FCC-ee study [3]. Because EMD2 focusing depends on final-state momenta, ISR and beamstrahlung can modify the quoted 1.4×10^-3 loss. In addition, the convergence test in Fig. 9 covers only the number of longitudinal slices; no check of transverse grid/time-step convergence or a comparison against another tracking code is reported. Please quantify these sensitivities or give a stronger justification for neglecting them.","section":"§2, §4, Fig. 9"},{"comment":"The quoted central values carry no statistical uncertainties. With samples of about 10^5 events and effects of 4×10^-3 and 1.4×10^-3, the statistical error on the extracted losses is at the level of 10^-4, comparable to the target precision and to the quoted systematic bounds. Reporting standard errors is necessary to assess whether the claimed correction uncertainties (5×10^-4 and 3×10^-4) are actually supported by the statistics.","section":"§3–4"}],"minor_comments":[{"comment":"'Reduction of the crossing angle of∼140 mrad' should read ∼140 µrad, consistent with Section 3.","section":"§6 Conclusion"},{"comment":"Typos: 'tme most sensitive' should be 'the most sensitive'; 'if, however, if one would count' is doubled.","section":"§3"},{"comment":"The figure captions and axes could be clearer about whether the plotted values are per-scan-point or integrated, and whether any statistical error bars are omitted.","section":"Figs. 7 and 11"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest about several limitations, but the central numbers rest on an unvalidated single-code simulation chain. The N=14 vs N=15 inconsistency and the lack of an independent cross-check are the main blockers. I would be willing to support acceptance after these are addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First things first: this is the first set of CEPC-specific estimates for initial- and final-state electromagnetic deflection effects on the low-angle Bhabha luminosity count, with post-CDR beam parameters. If the simulation is right, uncorrected losses are ~6×10^-3 — roughly sixty times the 10^-4 precision target — so the problem is real and the correction scheme is worth discussing. The paper is competent and straightforward: the GuineaPig+BHLUMI chain is described clearly, the internal consistency check (52 keV from E·tan(α/2)·Δα) checks out, and the EMD2 slice-count convergence test is a good idea. Credit is due for being honest that the effects have not been measured and that the simulation settings matter.\n\nNow the soft spots, in order of importance. First, there is no external validation of the GuineaPig kick or the boost/rotation association; a factor-of-two error in the kick would move the EMD1 correction by ~2–4×10^-3, far outside the goal. The ±10% beam-parameter scans in Figs. 7 and 11 only probe sensitivity around the central simulated point; they cannot validate the absolute scale. This is the same criticism the stress-test raises, and it lands. Second, the quoted residual systematics — 5×10^-4 for EMD1 and 3×10^-4 for EMD2 — are themselves larger than the 10^-4 target. Even at face value, the proposed corrections leave you above the precision budget. Third, there are small inconsistencies: the conclusion says 140 mrad where the body says 140 µrad; the Section 5 crossing-angle formula has a unit mismatch (260 mrad/√N cannot give ~1 µrad with 10^4–10^5 events); and the text mentions 2×10^-4 count variation but then quotes 5×10^-4 systematic. These look like typos, but they matter in a paper whose whole point is precision. Fourth, no error bars accompany the central values, and no code or configuration files are shipped, so the numbers are not reproducible as-is.\n\nNone of this breaks the central direction or order of magnitude. The paper is honest about its limitations and its method is standard for this area. It does not claim to have solved the 10^-4 problem; it quantifies an effect that must be dealt with and sketches a plausible correction path.\n\nWho is this for? People working on CEPC luminosity measurement and machine-detector interface. It deserves a serious referee: the numbers are new, the method is standard, and the caveats are identifiable. I would send it to peer review with a clear request for error bars, an independent check of the kick (even a rough analytic estimate), and fixing the unit/typo issues.","headline":"Useful first CEPC numbers for EMD effects on Bhabha counting, but the absolute scale hangs entirely on one simulation code and the residual uncertainties quoted are themselves 3–5× above the 10^-4 goal.","tokens_in":6893,"tokens_out":3949,"would_cite":true,"duration_ms":39230,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Electromagnetic fields of the CEPC bunches deflect low-angle Bhabha pairs enough to shift the integrated luminosity by about 6 times 10^-3 if unaccounted for, and the paper shows a simulation-based correction can reduce this to near the 10^","keywords":["CEPC","integrated luminosity","Bhabha scattering","electromagnetic deflection","beam-beam effects","crossing angle","luminometer","Z pole"],"falsifier":"Use the published Bhabha-count-loss versus crossing-angle dependence to predict the luminosity loss, then measure the crossing angle with di-muon data at about 1 micro-radian; if the corrected count does not flatten to the predicted residual level, the simulation-based correction is wrong. A direct test of the input: compute the EMD1 kick (about 5.8 MeV, 140 micro-radian crossing-angle reduction) with an independent beam-beam code or an analytic bunch-field integral; if a second estimate differs by about a factor of two, the correction error alone already exceeds the 10^-4 goal.","tokens_in":5750,"feed_emoji":"⚛️","tokens_out":6440,"duration_ms":68945,"temperature":0.7,"pith_summary":"At the CEPC Z pole, the electromagnetic fields of the oncoming bunches deflect both the initial electron and positron and the final-state Bhabha particles. The paper simulates these effects and finds that they destroy the back-to-back collinearity of the low-angle Bhabha pairs counted by the luminometer, so the raw luminosity count is low by about 6e-3, roughly 4e-3 from initial-state deflection and 1.4e-3 from final-state deflection. The authors show that simulation can supply the correction: under +/-10% beam-parameter variations the residual uncertainty stays at most 5e-4 for the initial-state effect and 3e-4 for the final-state effect, and a ~1 micro-radian measurement of the crossing angle from ~70 inverse-picobarns of di-muon data can anchor the dominant correction. If these numbers hold, the deflection can be corrected and the CEPC luminosity measurement can approach its 10^-4 precision goal.","feed_headline":"Bunch fields shift CEPC luminosity count by 0.6%","feed_subtitle":"Simulation can correct the beam-deflection loss and keep CEPC's Z-pole luminosity near its 10^-4 target.","key_machinery":"The central mechanism is the electromagnetic deflection of low-angle Bhabha particles in the fields of the oncoming bunches, split into an initial-state effect (EMD1) and a final-state effect (EMD2). The study generates roughly 600,000 Bhabha events, associates them event-by-event with the deflected initial state through a boost and rotation that makes the displaced initial system the center-of-mass frame of the final state, and tracks the final-state particles in the opposite-charge bunch fields. This yields two observable handles: the reduction of the effective crossing angle (about 140 micro-radians) and the acollinearity of the Bhabha pair (about 170 micro-radians for EMD1 and 43 micro-r","core_discovery":"The paper's central claim is that at CEPC with post-CDR beams, the electromagnetic fields of incoming bunches deflect initial-state particles by a mean kick of about 5.8 MeV to the e+e- system and reduce the effective crossing angle by about 140 micro-radians, and they deflect the low-angle Bhabha final states so their acollinearity changes by about 43 micro-radians. Propagating both effects through the luminometer count gives a relative loss of integrated luminosity of about 6e-3 if uncorrected: about 4e-3 from EMD1, caused by loss of collinearity at the inner aperture, and about 1.4e-3 from EMD2. The authors argue this loss is correctable from simulation, with stability under +/-10% bunch-","pith_inferences":["A natural cross-check not explored in the paper: the EMD1 kick also shifts each beam's energy by about 52 keV on average, so comparing that shift with the predicted crossing-angle reduction and with beam-energy measurements would test the simulation independently of Bhabha counting.","The same correction scheme should scale to other Z-pole circular colliders, where the count loss grows with bunch intensity and shrinks with beam energy; an explicit scaling law would make the CEPC numbers portable.","The predicted azimuth-dependent acollinearity could be measured in data with the tracking layer in front of the luminometer, giving an in-situ validation of the correction before it is applied.","If the EMD2 loss is confirmed by measuring acollinearity, it could double as a bunch-shape diagnostic, since the effect is sensitive to bunch-length variations."],"forward_implications":["Raw Bhabha counts at the CEPC Z pole carry a roughly 0.6% correction before the 10^-4 luminosity precision can be claimed.","The EMD1 part of the correction is anchored by a ~1 micro-radian measurement of the crossing angle from di-muon events, achievable with about 70 inverse picobarns of data.","Under +/-10% beam-parameter variations, simulation-based corrections leave residual luminosity uncertainties no larger than 5e-4 (EMD1) and 3e-4 (EMD2).","Counting asymmetrically in the two luminometer arms (55-77 mrad in one, 53-79 mrad in the other) would reduce the EMD1 loss to about 6e-5 for nominal beams, provided the detector sits on the outgoing-beam axis.","Because both effects grow toward the inner aperture, moving the luminometer fiducial volume inward trades counting statistics against larger correction uncertainties."],"fun_headline_variants":["CEPC luminosity loss from beam deflection: 0.6%, correctable","Simulation can fix CEPC's beam-deflection luminosity loss","Bunch fields cost CEPC 0.6% luminosity, correction simulated","Correcting the 0.6% luminosity dip from CEPC beam kicks","CEPC Z-pole luminosity: 0.6% beam-deflection error, fixable"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The correction chain assumes the beam-beam simulation faithfully computes the electromagnetic kick on the colliding particles and that translating that kick into Bhabha final states by event-by-event boost and rotation is accurate; no independent analytic estimate or second simulation code validates either step, and the paper itself flags sensitivity to simulation settings.","fun_headline_variants_meta":{"raw":{"variants":["CEPC luminosity loss from beam deflection: 0.6%, correctable","Simulation can fix CEPC's beam-deflection luminosity loss","Bunch fields cost CEPC 0.6% luminosity, correction simulated","Correcting the 0.6% luminosity dip from CEPC beam kicks","CEPC Z-pole luminosity: 0.6% beam-deflection error, fixable"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001117,"raw_usage":{"total_tokens":4461,"prompt_tokens":692,"completion_tokens":3769,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":436,"completion_tokens_details":{"reasoning_tokens":3664}},"tokens_in":436,"tokens_out":3769,"duration_ms":19238,"temperature":1.0,"reasoning_tokens":3664,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T00:28:31.177427+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Use the published Bhabha-count-loss versus crossing-angle dependence to predict the luminosity loss, then measure the crossing angle with di-muon data at about 1 micro-radian; if the corrected count does not flatten to the predicted residual level, the simulation-based correction is wrong. A direct test of the input: compute the EMD1 kick (about 5.8 MeV, 140 micro-radian crossing-angle reduction) with an independent beam-beam code or an analytic bunch-field integral; if a second estimate differs by about a factor of two, the correction error alone already exceeds the 10^-4 goal.","supporting_citations":[],"review_version":1}