{"id":"dc65f585-7f87-4e2a-9553-c03483848de1","arxiv_id":"2506.08453","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"KEDR measures B(J/ψ→ρη) = (2.04 ± 0.58 ± 0.39) x 10^-4 and B(J/ψ→φη) = (7.82 ± 1.17 ± 0.58) x 10^-4 from 4.93 million J/ψ decays, consistent with previous experiments.","lead":"The KEDR experiment used 4.93 million J/ψ decays to measure the branching fractions of J/ψ into ρη and φη final states, finding values consistent with earlier measurements. It also reports weak hints of contributions from the ρ(1450)η and a₂π intermediate states that a larger data sample could confirm or rule out.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Model dependence of the π+π−η Dalitz fit is the key risk: the 10.7% 'unaccounted interference' systematic is from a two-model comparison and may understate shifts in the ρη yield; a closure test against an a2π–ρη interference model is required.","rationale":"The reader's weakest assumption correctly identifies the Dalitz fit model as the load-bearing point. I agree, and the proposed closure test is the direct way to test whether the 10.7% systematic brackets the model uncertainty. The paper is honest and internally transparent: the systematic tables sum correctly, the 2.2% correction is applied consistently, and the central values agree with previous measurements. The p4/p5 notation slip in Eq. 13 is a concrete textual inconsistency, but it affects the derived B(π+π−η) rather than the headline ρη value, so it is secondary; it should nevertheless be corrected. The key risk remains that an unmodeled a2π–ρη interference could bias the fitted ρη yield beyond the quoted systematic. Since the reader already conditionalizes on this model dependence, my read does not change the verdict.","tokens_in":12993,"tokens_out":7437,"duration_ms":88420,"concrete_test":"Run a pseudo-experiment closure test: generate 1000 toy samples from the fitted Eq. 8 model plus an additional a2π–ρη interference term with relative phase scanned in [0,2π) and magnitude from the a2π fit uncertainty; fit each toy with the paper's Eq. 8 model and record the bias in p1 (ρη yield). If the RMS bias exceeds the quoted 10.7% systematic, the central B(J/ψ→ρη) uncertainty is underestimated; if it stays below, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The ρη branching fraction is extracted from the binned Dalitz fit of Eq. 8, in which the a2π term is added incoherently (Eq. 5) and only V+P resonances interfere. The paper's largest systematic, 10.7% (Sec. 6.2), is obtained by comparing two fit variants: interference with a2π fixed at zero phase versus no interference. This does not cover the possibility that a2π interferes with ρη at an unknown phase with magnitude comparable to the fitted a2π yield (23/134 events); such an interference could shift the ρη yield p1 and hence B(J/ψ→ρη) by more than the quoted 10.7%. The 17% selection efficiency is also computed from the same generator, so a wrong interference model enters both the fitted yield and the efficiency. With only 67 ρη events the two-model comparison has limited statistical power. No data or code release is provided, so the closure of the fit model cannot be checked externally. A separate internal inconsistency: Eq. 13 labels the π+π−π0γγ subtraction as p4 while Eq. 8 uses p5 for that component; this affects the derived B(J/ψ→π+π−η) if literal.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports measurements of the branching fractions B(J/ψ→ρη) and B(J/ψ→φη) using 4.93 million J/ψ events collected with the KEDR detector. The ρη result is obtained from a binned Dalitz-plot fit to J/ψ→π+π−η, where the signal is modeled as a coherent sum of ρη, ωη, and ρ(1450)η amplitudes with an incoherent a2±π∓ term and fixed continuum and background contributions. The φη result is obtained from a one-dimensional fit to the K+K− invariant mass in J/ψ→K+K−η. The paper also reports B(J/ψ→π+π−η), B(J/ψ→a2±π∓), an upper limit on B(J/ψ→ρ(1450)η→π+π−η), and the ρ−ω and ρ−ρ(1450) interference phases. All quoted numerical results are internally arithmetically consistent: the 2.2% correction reconciles the Section 6.1 values with the Section 7 values, and the systematic tables add in quadrature to the stated totals.","tokens_in":13226,"tokens_out":6031,"duration_ms":80530,"significance":"If the results hold, the paper provides an independent KEDR confirmation of the world-average values for B(J/ψ→ρη) and B(J/ψ→φη) from a dataset that is modest in size but complementary to BES/BaBar. The dynamic analysis of π+π−η adds useful, albeit low-significance, information on possible ρ(1450)η and a2±π∓ contributions. The paper is careful to quote separate statistical and systematic uncertainties, uses external normalization anchors (PDG branching fractions, BES-II B(J/ψ→ωη), BaBar continuum cross-section), and the internal arithmetic is sound. The main limitation is that the largest systematic uncertainty in the ρη measurement, the 10.7% assigned to unaccounted interference effects, is estimated from a two-model comparison rather than from a data-driven or closure-test constraint, and the same model enters the efficiency calculation. This prevents the central ρη result from being as robust as the quoted uncertainty suggests.","major_comments":[{"comment":"The largest systematic uncertainty in the ρη measurement, 10.7% from 'unaccounted interference effects between tensor mesons and between V+P and T+P processes,' is estimated by comparing two fit variants: one in which interference with a2±π∓ is included at zero phase and one in which it is excluded. This comparison does not scan the interference phase or its magnitude. Given that the fitted a2±π∓ yield is 23 events out of 134 and that p1 (the ρη strength) is determined in the same fit, an a2π–ρη interference term at an arbitrary phase and with a magnitude comparable to the fitted a2π yield could shift p1 by more than this estimate. In addition, the 17% selection efficiency for ρη (Section 5.1) is computed with the same generator model, so an incorrect interference model enters both the fitted yield and the efficiency. A closure test with pseudo-experiments generated from a model with a2π–ρη interference at several phases and magnitudes, then fitted with the nominal model, should be performed (or a data-driven constraint on this interference should be provided) before the 10.7% systematic can be considered to cover the dominant model uncertainty.","section":"Section 6.2 and Eq. (8)"},{"comment":"There is an internal inconsistency in the definition of the π+π−π0γγ background subtraction. Equation (8) uses p5 as the strength of the Hπ+π−π0γγ histogram, while Eq. (13) defines Nπ+π−π0γγ = p4 Iπ+π−π0γγ, where p4 is the a2±π∓ strength in Eq. (8). If taken literally, the subtraction used to obtain Nπ+π−η in Eq. (13) uses the wrong fit parameter, which directly affects the derived B(J/ψ→π+π−η) in Eq. (10). The authors should correct this notation and verify explicitly that the quoted B(J/ψ→π+π−η) value is unaffected.","section":"Section 5.1, Eqs. (8) and (13)"},{"comment":"The paper reports 'hints' of ρ(1450)η and a2±π∓ contributions, but no statistical significance is given for the improvement of the fit when these components are added. The statement that the fit without these intermediate states gives 'worse likelihood function values' should be quantified with Δ(-2lnL) or a p-value, and the procedure used to set the 90% confidence-level upper limit on ρ(1450)η should be described (e.g., profile likelihood or Feldman-Cousins). This is needed to support the secondary dynamical claims, especially because the extracted a2±π∓ branching fraction has a central value only about two standard deviations from zero.","section":"Section 6.1 and Section 7"}],"minor_comments":[{"comment":"The text refers to 'Belle-II' in the list of experiments with consistent results, but reference [7] and the figure label are for Belle. Please correct this inconsistency.","section":"Section 7 and Fig. 5"},{"comment":"The resonance amplitude in Eq. (2) is written with an explicit B(J/ψ→ρη)B(ρ→ππ) factor, but the relation between the fit parameter p1 and this branching fraction is not shown explicitly. It would improve clarity to state the proportionality constant and how it is absorbed into Eq. (10).","section":"Section 2.1, Eq. (2)"},{"comment":"The Monte Carlo generators from references [13] and [14] are internal to the KEDR collaboration. Please provide a brief description of the assumed production angular distributions and any version/parameter choices, so that the interference-term histograms Hc± and Hs± are reproducible from the text.","section":"Section 3"},{"comment":"The 2.2% upward correction from track inefficiency and photon loss is applied to all branching fractions, but this correction is not listed in Tables 3 or 5. Please state clearly in the table captions or text that the quoted central values include this correction, and list the 2.2% as a separate item for transparency.","section":"Section 6.2"},{"comment":"The non-resonant K+K−η contribution is simulated with an 'infinite decay width' of the vector particle (Section 3). This is a potentially ambiguous line shape; please specify the functional form used and how the 4.3% systematic was obtained from the alternative phase-space simulation.","section":"Section 6.4"}],"recommendation":"major_revision","confidential_remarks":"The paper's internal arithmetic and external normalization anchors are sound, and the φη result is relatively straightforward. The central issue is the ρη result: the largest systematic is an estimate from a two-model comparison, and the same model uncertainty enters the efficiency. I recommend requesting closure tests and the fix of the p4/p5 notation before publication. The paper is within the scope of Nuclear Physics B."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Fairly straightforward: this is a competent, transparent measurement paper from KEDR that confirms two known branching fractions with an independent detector and adds some weak dynamical hints. The headline numbers look believable, but the model-dependence of the ρη extraction is the main risk, and there is a small internal inconsistency in Sec. 5 that should be fixed.\n\nWhat's actually new: this is the first KEDR measurement of J/ψ→ρη and φη from their 4.93M-event sample. The π+π−η Dalitz fit that includes ρ(1450)η, a2π, and interference phases goes beyond what BaBar and BES-III did for this channel. The paper does a good job of reporting its systematics honestly—the 10.7% 'unaccounted interference' term is the largest and is described as a model-comparison estimate, not hidden. The internal arithmetic checks out: the 2.2% correction is applied consistently, the systematic tables sum correctly to 19.3% and 7.4%, and the event counts line up with the quoted efficiencies and branching fractions.\n\nThe soft spots are real but not fatal. The 10.7% systematic comes from comparing two fit models (interference with a2π at zero phase vs. no interference). That does not cover an arbitrary a2π–ρη interference phase or additional ρ' states, and with only 67 signal events the test has limited power. So the quoted uncertainty on B(ρη) could understate the model sensitivity. Also, there is no data or code release, so external closure checks are impossible. The a2π branch has only about 2σ combined significance, so calling it a measured branching fraction in the abstract is a bit generous; a significance statement would be better. And the p4/p5 mix-up: Eq. 8 uses p5 for π+π−π0γγ, but Eq. 13 says p4. That looks like a typo, but as written it is an inconsistency in the definition of Nπ+π−η.\n\nThe citation pattern is fine: the self-citations point to the collaboration's own generators and corrections, and the normalization anchors are external PDG/BES-II values.\n\nBottom line: this is a solid confirmation measurement with some underpowered dynamical hints. It deserves a serious referee, mainly to ask for the model-dependence to be addressed more thoroughly and the typo fixed. I wouldn't block publication over the model issue—the paper is transparent about it—but the referee should push on whether the 10.7% truly covers the variation. Worth a reading-group discussion if you care about how Dalitz-fit systematics are estimated.","headline":"A competent, transparent KEDR measurement that independently confirms the ρη and φη branching fractions and adds weak dynamical hints; the main risk is model-dependence of the ρη extraction, plus a small internal inconsistency in Eq. 13.","tokens_in":14765,"tokens_out":2503,"would_cite":false,"duration_ms":28706,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["13.25.Gv"],"model":"deepseek-v4-flash","headline":"Using 4.93 million J/ψ events, the authors measure B(J/ψ→ρη) = (2.04 ± 0.58 ± 0.39) × 10⁻⁴ and B(J/ψ→φη) = (7.82 ± 1.17 ± 0.58) × 10⁻⁴, and find hints of ρ(1450)η and a₂±π∓ contributions in the π⁺π⁻η Dalitz plot.","keywords":["J/ψ decays","branching fraction","ρ η final state","φ η final state","Dalitz plot analysis","charmonium","e+e− experiments","intermediate states"],"falsifier":"A model-independent partial-wave analysis of J/ψ→π⁺π⁻η on a high-statistics sample would settle the issue: if it does not reproduce the ρη yield of 67 events and the ρ(1450)η upper limit of 1.51×10⁻⁴, the central branching fraction would need to be revised.","tokens_in":12633,"feed_emoji":"⚛️","tokens_out":16521,"duration_ms":158734,"temperature":0.7,"pith_summary":"The paper reports new measurements of the branching fractions of the J/ψ meson into ρη and φη final states, based on 4.93 million J/ψ events collected with the KEDR detector. The results—B(J/ψ→ρη)=(2.04±0.58±0.39)×10⁻⁴ and B(J/ψ→φη)=(7.82±1.17±0.58)×10⁻⁴—agree with earlier experiments, providing an independent check of the world-average values. For the ρη channel, a fit to the π⁺π⁻η Dalitz plot finds hints of contributions from ρ(1450)η and a₂±π∓ intermediate states, with a 90% confidence upper limit on the ρ(1450)η contribution. If correct, the measurements confirm the standard picture of these charmonium decays while suggesting that the π⁺π⁻η final state contains more dynamics than a simple ρη plus ωη description.","feed_headline":"Measured: J/ψ→ρη=2.04×10⁻⁴, φη=7.82×10⁻⁴","feed_subtitle":"An independent 4.93M-event sample confirms the world-average J/ψ decay rates and reveals possible new intermediate states.","key_machinery":"The central object is the Dalitz plot of J/ψ→π⁺π⁻η, modeled as the coherent sum of vector–pseudoscalar amplitudes (ρη, ωη, ρ(1450)η) with energy-dependent line shapes and relative phases, plus an incoherent tensor–pseudoscalar a₂±π∓ term. Each resonance amplitude uses a Breit–Wigner form with an energy-dependent width (Eqs. 2–3), and the interference terms are included in cosine and sine parts with Monte Carlo templates. The fit minimizes a binned Baker–Cousins likelihood (Eq. 9). For φη, the machinery reduces to a one-dimensional fit of the K⁺K⁻ invariant mass with a ϕ line shape plus a non-resonant contribution. The same generator (from the earlier KEDR analysis) provides the efficiency corrections and background templates.","core_discovery":"The central claim is that B(J/ψ→ρη)=(2.04±0.58±0.39)×10⁻⁴ and B(J/ψ→φη)=(7.82±1.17±0.58)×10⁻⁴, extracted from 4.93 million J/ψ decays recorded at the KEDR detector. The ρη value comes from a fit to the π⁺π⁻η Dalitz plot that models the final state as coherent ρη, ωη, and ρ(1450)η amplitudes plus an incoherent a₂±π∓ term; the fit yields 67 ρη events out of 134 total and the additional branching fractions B(J/ψ→π⁺π⁻η)=(4.73±0.49±1.17)×10⁻⁴, B(J/ψ→a₂±π∓)=(1.05±0.37±0.34)×10⁻³, and B(J/ψ→ρ(1450)η→π⁺π⁻η)<1.51×10⁻⁴ at 90% CL. The φη result is obtained from a fit to the K⁺K⁻ invariant mass with a ϕ resonance and a non-resonant term. The paper reports that all results are consistent with previous experiments and provides interference phases between ρ–ω and ρ–ρ(1450) as model inputs.","pith_inferences":["If the a₂±π∓ hint is real, its branching fraction of about 1.05×10⁻³ falls well below the theoretical estimate of 3.8×10⁻³ cited in the paper, suggesting the production mechanism may suppress this mode or the estimate is too high.","A high-statistics sample from a future experiment could turn the ρ(1450)η hint into a signal or push the upper limit below 1.51×10⁻⁴, directly testing the model used here.","The dominant 10.7% systematic from unaccounted interference is a model-comparison estimate; a data-driven extraction of the interference phases from angular distributions would reduce this uncertainty.","The 17% efficiency for ρη is computed with the same generator model used in the fit; an independent efficiency calibration would verify the absolute branching-fraction scale."],"forward_implications":["The measured branching fractions independently confirm the world-average values for J/ψ→ρη and J/ψ→φη, strengthening the empirical basis for these decay rates.","The Dalitz fit implies that J/ψ→π⁺π⁻η is not fully described by ρη and ωη alone; the contributions from ρ(1450)η and a₂±π∓ found in the model should be included in future amplitude analyses.","The interference phases between ρ–ω and ρ–ρ(1450) provide new constraints on the relative strong and electromagnetic amplitudes in J/ψ decays into vector–pseudoscalar pairs.","The 90% CL upper limit on ρ(1450)η→π⁺π⁻η bounds the excited-vector contribution, complementing studies of the analogous π⁺π⁻π⁰ channel."],"supporting_citations":[{"why":"Supplies the MARK-III measurement of B(J/ψ→ρη) that the new result is compared with.","marker":"[1]"},{"why":"Supplies DM2 measurements of J/ψ→ρη and φη used for consistency comparison.","marker":"[2]"},{"why":"Provides the BaBar cross section for e+e−→π+π−η used to fix the continuum background.","marker":"[3]"},{"why":"Provides the BES-III study of J/ψ→π+π−η that the Dalitz analysis extends.","marker":"[4]"},{"why":"Supplies the BES-II measurement of B(J/ψ→φη) used for comparison.","marker":"[8]"},{"why":"Defines the binned likelihood (Baker–Cousins) used for all fits in this paper.","marker":"[12]"},{"why":"Supplies the vector–pseudoscalar event generator, background generators, and the 2.2% efficiency correction used in both channels.","marker":"[13]"},{"why":"Provides the particle properties and branching fractions (η→γγ, a2→πη, ϕ→K+K−) used as inputs.","marker":"[17]"},{"why":"Supplies the B(J/ψ→ωη) constraint from BES-II that fixes the ωη contribution in the ρη fit.","marker":"[21]"}],"fun_headline_variants":["J/ψ→ρη=2.04e-4, φη=7.82e-4 from KEDR","KEDR reports J/ψ decay rates to ρη and φη","Possible ρ(1450) and a2π contributions in J/ψ→ρη","4.93M J/ψ events refine ρη and φη branching fractions","New J/ψ→ρη and φη branching fractions measured"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the assumed interference model for the π⁺π⁻η Dalitz plot—coherent ρη, ωη, ρ(1450)η plus incoherent a₂±π∓ with the specified line shapes; if additional interference structure exists, the ρη yield and the 17% efficiency correction could shift beyond the quoted uncertainties.","fun_headline_variants_meta":{"raw":{"variants":["J/ψ→ρη=2.04e-4, φη=7.82e-4 from KEDR","KEDR reports J/ψ decay rates to ρη and φη","Possible ρ(1450) and a2π contributions in J/ψ→ρη","4.93M J/ψ events refine ρη and φη branching fractions","New J/ψ→ρη and φη branching fractions measured"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001064,"raw_usage":{"total_tokens":4616,"prompt_tokens":1254,"completion_tokens":3362,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":870,"completion_tokens_details":{"reasoning_tokens":3250}},"tokens_in":870,"tokens_out":3362,"duration_ms":29491,"temperature":1.0,"reasoning_tokens":3250,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:13:25.351578+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A model-independent partial-wave analysis of J/ψ→π⁺π⁻η on a high-statistics sample would settle the issue: if it does not reproduce the ρη yield of 67 events and the ρ(1450)η upper limit of 1.51×10⁻⁴, the central branching fraction would need to be revised.","supporting_citations":[{"cited_title":"Coffman et al","cited_arxiv_id":null,"evidence_quote":"Supplies the MARK-III measurement of B(J/ψ→ρη) that the new result is compared with."},{"cited_title":"Jousset et al","cited_arxiv_id":null,"evidence_quote":"Supplies DM2 measurements of J/ψ→ρη and φη used for consistency comparison."},{"cited_title":"Lees et al","cited_arxiv_id":null,"evidence_quote":"Provides the BaBar cross section for e+e−→π+π−η used to fix the continuum background."},{"cited_title":"Ablikim et al","cited_arxiv_id":null,"evidence_quote":"Provides the BES-III study of J/ψ→π+π−η that the Dalitz analysis extends."},{"cited_title":"Ablikim et al","cited_arxiv_id":null,"evidence_quote":"Supplies the BES-II measurement of B(J/ψ→φη) used for comparison."},{"cited_title":"Baker, R.D","cited_arxiv_id":null,"evidence_quote":"Defines the binned likelihood (Baker–Cousins) used for all fits in this paper."},{"cited_title":"Anashin et al., (KEDR Collaboration), Measurement of the branch- ing fraction of J/ψ → ρπ at KEDR , JHEP 06 (2023) 196","cited_arxiv_id":null,"evidence_quote":"Supplies the vector–pseudoscalar event generator, background generators, and the 2.2% efficiency correction used in both channels."},{"cited_title":"Navas et al.), Review of particle physics , Phys","cited_arxiv_id":null,"evidence_quote":"Provides the particle properties and branching fractions (η→γγ, a2→πη, ϕ→K+K−) used as inputs."},{"cited_title":"Ablikim et al","cited_arxiv_id":null,"evidence_quote":"Supplies the B(J/ψ→ωη) constraint from BES-II that fixes the ωη contribution in the ρη fit."}],"review_version":1}