{"id":"4a2acf9b-1736-445c-aa16-0cfd0ad29c48","arxiv_id":"2607.20366","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The suppression of X(2370)→K*Kbar indicates X(2370 is a flavor-singlet, supporting the interpretation that the lightest 0−+ glueball is its dominant constituent.","lead":"Using 10 billion J/ψ decays, BESIII finds no sign of X(2370) decaying to K*(892) and K mesons, setting an upper limit that points to the particle being flavor-neutral. The result adds weight to the long-standing idea that X(2370) is mostly made of glue—the lightest predicted 'glueball' of QCD.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Partial-width bound depends on an unvalidated production-rate estimate; without it, the η–η' excitation exclusion weakens.","rationale":"The paper's experimental upper limit is a solid, self-contained result. The leap from it to 'dominant glueball constituent' is a synthesis that depends on several intermediate steps: the G-parity suppression being exact for a flavor-singlet, the production-rate floor, the comparison with η–η' excitation expectations, and the consistency of the full property list. The single most load-bearing step is the conversion of the product-BF limit into a partial-width limit using B[J/ψ→γX] > 1e−3, because that conversion is what quantitatively excludes the strongest alternative (η–η' excitation). The reader identified this as a second load-bearing premise; I agree that it is the weakest link. The G-parity argument is theoretically standard for a pure flavor-singlet decay, though dynamical suppression could mimic it; the production-rate floor, by contrast, has no direct empirical anchor in the paper and is explicitly labeled a 'reasonable estimate.' The paper's own cited model [21] contradicts the measured decay pattern, so the estimate cannot lean on it. If the production rate is lower, the K*Kbar bound weakens and the discrimination against the η–η' excitation picture loses its quantitative force, leaving the glueball conclusion plausible but unsupported at the claimed level. I therefore recommend keeping the CONDITIONAL verdict rather than upgrading to ACCEPT; the experimental result stands, but the central theoretical claim warrants an explicit check of the production-rate assumption. This is not a fatal flaw and no dishonesty is implied — the paper is transparent about the estimate — but it is the place where the argument is least secure.","tokens_in":19208,"tokens_out":1695,"duration_ms":13730,"concrete_test":"Replace the asserted floor B[J/ψ→γX] > 1e−3 with the largest estimate consistent with existing BESIII BF measurements: sum the measured product BFs in Table I (≈8.8e−4) and add any unobserved modes bounded by data; if the resulting B[J/ψ→γX] can be as low as ~3e−4, recompute the K*Kbar partial-width limit and check whether Γ < 2 MeV survives. Alternatively, perform an amplitude analysis of J/ψ→γX in the X mass region to extract B[J/ψ→γX] directly from the KsKsπ0 and π0π0η spectra without fixing the X width, and see whether the value exceeds 1e−3.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that X(2370) is the dominant 0−+ glueball rests on excluding the main competing interpretation, the η–η' radial excitation, via the bound Γ[X→K*Kbar] < 2 MeV. That bound is derived in the paper as B[J/ψ→γX] × B[X→K*Kbar] < 2.7e−6 divided by the 'reasonable estimate' B[J/ψ→γX] > 1e−3. This estimate is asserted without a cited measurement or derivation, and it is load-bearing: if the true production rate is, say, 3e−4, the partial-width bound loosens by more than a factor of three and the claimed exclusion of the 15–200 MeV expectation fails. The paper itself notes the only quoted model calculation [21] gives B[J/ψ→γX] > 2.87e−3, but flags that same model as inconsistent with measured decay BFs; other model rates may be lower. The G-parity flavor-singlet argument is stronger but only excludes a pure qqbar nonet state; it does not by itself rule out an η–η' excitation with a singlet component. So the load-bearing quantitative step is the unvalidated production-rate floor.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a search for the decay X(2370)->K*(892)bar{K}+c.c. using the process J/psi->gamma K_S^0 K_S^0 pi^0 in 10 billion J/psi events at BESIII. No signal is observed; the product branching fraction B(J/psi->gamma X) x B(X->K*bar{K}->K_S^0 K_S^0 pi^0) is <2.7e-6 at 90% CL, and the ratio to B(X->K_S^0 K_S^0 pi^0) is R<0.081. The authors interpret the suppression of the K*bar{K} mode, together with a production-rate estimate B(J/psi->gamma X)>1e-3, as excluding the eta-eta' excitation interpretation and supporting a flavor-singlet nature. They summarize all BESIII measurements of X(2370) and argue that a dominant component of the lightest 0^-+ glueball is essential to explain the mass, J^PC, production rate, decay pattern, flavor-singlet property, narrow partial width, and suppressed radiative decays to omega/phi, concluding that X(2370) is predominantly the lightest 0^-+ glueball.","tokens_in":19507,"tokens_out":6956,"duration_ms":52575,"significance":"The new upper limit is a valuable experimental result from the BESIII data set, and the analysis follows standard procedures. If the flavor-singlet interpretation is correct, this would be an important step toward identifying the first glueball. The paper's strength is its comprehensive aggregation of BESIII measurements and the use of a model-independent selection rule (generalized G-parity) to argue for flavor-singlet behavior. However, the central quantitative claim rests on an unvalidated production-rate floor and on an idealized interpretation of the selection rule, so the conclusion as stated is not yet fully established.","major_comments":[{"comment":"The bound Γ[X->K*bar{K}]<2 MeV is obtained by dividing the product-BF upper limit by the 'reasonable estimate' B[J/psi->gamma X]>1e-3. This estimate is asserted without derivation or uncertainty. The sum of the four measured product BFs in Table I is 8.78e-4, so a lower bound of 1e-3 does not follow from these data alone. If the true production rate is, e.g., 3e-4, the bound becomes ~6.7 MeV, still below the 15 MeV lower edge of the eta-eta' expectation; but for 1e-4 it becomes ~20 MeV, overlapping the expected 15-200 MeV range. The quantitative exclusion of the eta-eta' excitation therefore depends critically on an unsubstantiated external input. Please provide a derivation or a conservative value for the production-rate floor, or soften the exclusion claim.","section":"Text after Table I (partial-width bound)"},{"comment":"The suppression of K*bar{K} is interpreted as proof of flavor-singlet nature via generalized G-parity. This selection rule is exact only for a pure 0^-+ flavor-singlet. A state with a small octet or q-qbar component would also have a suppressed but non-zero K*bar{K} width, so the upper limit constrains but does not eliminate such admixtures. In particular, an eta-eta' excitation with a predominantly singlet wavefunction could satisfy this limit. The claim that X(2370) is 'inconsistent with the eta-eta' excitation interpretation' is therefore stronger than what the data and the selection rule warrant. A quantitative bound on the allowed octet admixture or a discussion of singlet-dominated eta-eta' states is needed.","section":"Flavor-singlet / generalized G-parity (paragraph after Fig. 2)"},{"comment":"This statement is not obviously correct. Several established mesons above 1 GeV, such as f0(1370), f0(1500), and f0(1710), are commonly assigned large flavor-singlet components (and have been discussed as glueball candidates). The authors should qualify the claim, e.g., 'the first light meson above 1 GeV for which flavor-singlet nature is established through the generalized-G-parity selection rule,' or provide a citation substantiating uniqueness.","section":"Paragraph: 'It is the first flavor-singlet light hadron observed above 1 GeV/c^2'"},{"comment":"The concluding claim that a dominant glueball component is 'essential for a natural and complete explanation' is stronger than the evidence assembled. Each individual property is consistent with the glueball hypothesis, but the paper does not provide a quantitative measure of the glueball fraction or a rigorous exclusion of all alternatives. The production-rate estimate and the model-dependent expectations for eta-eta' excitations do not exhaust the possible non-glueball interpretations (e.g., tetraquarks or hybrids with singlet components). Suggest rewording to 'disfavors other interpretations' rather than 'essential for a natural and complete explanation.'","section":"Summary / conclusion"}],"minor_comments":[{"comment":"The word 'Gassian' should be 'Gaussian.'","section":"Fit description, text after Fig. 2"},{"comment":"The shorthand B_{K*bar{K}} is used without being explicitly defined as B[J/psi->gamma X] * B[X->K*(892)bar{K}] for the charged-conjugate-inclusive final state. Please define it at first use.","section":"Eq. (1) and surrounding text"},{"comment":"PHSP is used in figure captions without spelling out 'phase space.' Consider defining at first occurrence.","section":"Figure captions"},{"comment":"The isospin factors are listed but the procedure for applying them is not described. A brief explanation of how the factors 12, 3, 1.5, 4, and 2 are obtained would help the reader.","section":"Table I caption"}],"recommendation":"major_revision","confidential_remarks":"The experimental upper limit itself appears solid and is a useful new constraint. The main weakness is that the interpretation section over-relies on the 'reasonable estimate' B(J/psi->gamma X)>1e-3, which is not derived, and on a selection rule being exact for a pure singlet while the object under study is itself a mixture. The authors may be able to address this by softening the language and by being more careful about the production-rate floor, which would make the paper suitable for publication. I recommend major revision to allow them to tighten the logic."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about this one. First, the new experimental result is solid: BESIII searches for X(2370)→K*(892)Kbar in 10 billion J/ψ events, sees nothing, and sets B[J/ψ→γX]×B[X→K*Kbar→K_S K_S π^0] < 2.7×10^-6 at 90% CL, with R<0.081. That upper limit, plus the generalized-G-parity argument, makes a decent case that X(2370) is a flavor-singlet — the first such light hadron above 1 GeV. Second, the paper's bigger claim — that this makes the lightest 0^-+ glueball the dominant constituent — rests on much softer ground, and the title overstates it.\n\nThe analysis is careful. The signal extraction, background description, and systematics (parameterization, PHSP, additional resonances) are standard and described well enough to trust the upper limit. The paper also compiles all BESIII measurements of X(2370) into one table, which is genuinely useful.\n\nWhere it gets shaky: to go from the product upper limit to a partial-width limit of <2 MeV, they divide by an assumed B[J/ψ→γX(2370)] > 1×10^-3, described as a 'reasonable estimate.' No derivation, no cited measurement. The only model calculation they quote gives >2.87×10^-3 but is inconsistent with their measured branching fractions; other models might give lower rates. If the true production rate is, say, 3×10^-4, the partial-width bound loosens by more than a factor of three and the claimed exclusion of the 15–200 MeV expectation for an η-η' excitation no longer works. The authors flag the inconsistency but don't explore its consequences.\n\nThe G-parity flavor-singlet argument is the stronger piece, but even it has a caveat: it forbids K*Kbar for a pure 0^-+ flavor-singlet, so it excludes a pure quark nonet state. An η-η' excitation with a singlet component could, in principle, get around that. The paper doesn't address that possibility directly.\n\nThat said, the paper is honest about what is a consistency check and what is measurement. The upper limit is a genuinely new experimental result; the glueball interpretation is a plausible synthesis, not a proven identification. I'd send this to a referee — the experimental content deserves publication, and the interpretation is exactly what referees are for. If it goes through, it will be a useful reference for the X(2370) and next steps (ωω, φφ, K*(1410)Kbar searches).","headline":"Solid new upper limit on X(2370)→K*Kbar; the flavor-singlet case is decent, but the 'dominant glueball' claim leans on an unvalidated production-rate estimate.","tokens_in":19998,"tokens_out":4169,"would_cite":true,"duration_ms":33029,"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":"The X(2370) meson's every measured property — mass, spin-parity, production rate, decay pattern, narrow widths, and now flavor-singlet status — points to the lightest 0−+ glueball as its dominant constituent.","keywords":["X(2370)","pseudoscalar glueball","flavor-singlet hadron","generalized G-parity","J/ψ radiative decay","K*(892)Kbar suppression","OZI rule","lattice QCD glueball mass"],"falsifier":"Observing X(2370)→K*(892)0 Kbar0 at a product branching fraction above 2.7×10−6 (or R>0.081) would break the flavor-singlet identification; alternatively, a direct measurement of B[J/ψ→γX(2370)] below 1×10−3 would remove the quantitative exclusion of the η-η′ excitation. Both tests are within reach of the existing 10-billion-event J/ψ sample.","tokens_in":19100,"feed_emoji":"⚛️","tokens_out":10831,"duration_ms":75999,"temperature":0.7,"pith_summary":"This paper argues that the X(2370) particle is not an ordinary quark state but is dominated by the lightest 0−+ glueball, a bound state of gluons predicted by QCD. The new evidence is a search for X(2370)→K*(892)0 Kbar0: none is found, with a 90%-confidence upper limit of 2.7×10−6 on the product branching fraction. Because a 0−+ flavor-singlet meson is forbidden by generalized G-parity to decay into K*(892)0 Kbar0, the suppression identifies X(2370) as flavor-singlet — a first for any light hadron above 1 GeV/c². Assembled with earlier measurements of its mass, spin-parity 0−+, high production rate in J/ψ radiative decays, decay pattern similar to η_c, narrow partial widths, and suppressed γω and γϕ modes, the paper concludes that these properties are all consistent with the lightest 0−+ glueball and that other interpretations are disfavored.","feed_headline":"X(2370) is a flavor singlet, passing a key glueball test","feed_subtitle":"The missing K*(892)K mode confirms X(2370) is flavor-singlet, as a 0−+ glueball must be.","key_machinery":"The load-bearing object is X(2370), a hadron at 2359 MeV/c² with J^PC=0−+, seen in J/ψ→γKS KS η′, γKS KS π0, γπ0π0η, and γπ+π−η′. The decisive new tool is the generalized G-parity selection rule: for a pure 0−+ flavor-singlet meson, decay into K*(892)0 Kbar0 plus conjugate is forbidden. Searching this mode in the 10-billion-event J/ψ sample therefore acts as an unambiguous flavor-singlet test, converting a null result into a positive identification. Secondary machinery includes the √OZI estimate relating glueball partial widths to OZI-allowed and η_c widths, and the product-branching-fraction upper limit combined with a production-rate estimate to bound Γ(X→K*K)<2 MeV.","core_discovery":"The central claim: the lightest 0−+ glueball is the dominant constituent of X(2370). The decisive new result is that X(2370) does not decay into K*(892)0 Kbar0, with B[J/ψ→γX(2370)]×B[X(2370)→K*K→KSKSπ0]<2.7×10−6 and R<0.081 at 90% confidence. Generalized G-parity forbids this decay for a 0−+ flavor-singlet, so the absence marks X(2370) as flavor-singlet. Combined with the measured mass (consistent with lattice QCD's 2.3–3.0 GeV/c² pseudoscalar glueball), spin-parity 0−+, high J/ψ radiative production, narrow partial widths (<2 MeV for K*K), η_c-like decay modes, and suppressed γω/γϕ, the paper argues the glueball interpretation is the only one explaining all properties. It also excludes the","pith_inferences":["If the glueball identification holds, the same strategy — using generalized G-parity-forbidden channels as flavor-singlet filters — could be applied to other glueball candidates, including the scalar and tensor sectors.","The tiny 2°–5° mixing angle invoked to explain the high J/ψ radiative production implies a small charmonium admixture in X(2370); this could be probed through its two-photon coupling and charmonium-like radiative transitions, which the paper does not address.","A direct measurement of B[J/ψ→γX(2370)] — rather than the estimated >1×10−3 — would turn the <2 MeV partial-width bound into a firm number and may be achievable by summing all observed decay modes in the existing sample.","If confirmed, the X(2370) decay pattern becomes a benchmark for testing OZI suppression and the √OZI estimate of glueball partial widths at a mass where phase space differs sharply from η_c."],"forward_implications":["If X(2370) is dominated by the 0−+ glueball, it is the first identified glueball and direct evidence that gluons bind into matter, a distinctive prediction of non-Abelian QCD.","It would anchor the lattice-QCD prediction of the lightest pseudoscalar glueball near 2.3–3.0 GeV/c² and calibrate the glueball–charmonium mixing mechanism.","The flavor-singlet identification predicts that X(2370) should decay flavor-symmetrically to ωω and φφ, while ωφ and K*(1410)Kbar should be suppressed or forbidden; these are testable with the same data.","The exclusion of the η-η′ excitation interpretation sharpens the search for non-glueball candidates and constrains the qqbar content of the state.","A precise determination of the qqbar admixture would follow from partial-wave analyses of interference with neighboring resonances X(2120), X(2260), and X(2600) in multiple final states."],"fun_headline_variants":["X(2370) no K*K decay: flavor singlet","Missing K*K decay pins X(2370) as glueball","X(2370) resists K*K, confirms flavor singlet","Glueball win: X(2370) skips K*K decay","X(2370) passes flavor-singlet test for glueball"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that generalized G-parity conservation is exact for a pure 0−+ flavor-singlet — so no other dynamical effect such as phase space, form factors, or nodes is hiding the missing K*(892)K mode — and that the J/ψ→γX(2370) production rate is really above 1×10−3, which is what converts the product upper limit into a <2 MeV partial width.","fun_headline_variants_meta":{"raw":{"variants":["X(2370) no K*K decay: flavor singlet","Missing K*K decay pins X(2370) as glueball","X(2370) resists K*K, confirms flavor singlet","Glueball win: X(2370) skips K*K decay","X(2370) passes flavor-singlet test for glueball"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00157,"raw_usage":{"total_tokens":6225,"prompt_tokens":985,"completion_tokens":5240,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":729,"completion_tokens_details":{"reasoning_tokens":5141}},"tokens_in":729,"tokens_out":5240,"duration_ms":30336,"temperature":1.0,"reasoning_tokens":5141,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T10:01:29.038410+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observing X(2370)→K*(892)0 Kbar0 at a product branching fraction above 2.7×10−6 (or R>0.081) would break the flavor-singlet identification; alternatively, a direct measurement of B[J/ψ→γX(2370)] below 1×10−3 would remove the quantitative exclusion of the η-η′ excitation. Both tests are within reach of the existing 10-billion-event J/ψ sample.","supporting_citations":[],"review_version":1}