{"id":"e37e9c31-5100-4168-97d7-9f1aa06639b2","arxiv_id":"1908.02807","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"If the X(3872) is a weakly bound charm-meson molecule, its J/psi pi+ pi- branching fraction lies between roughly 4% and 33%, and is likely several times larger than the measured resonance-feature fraction.","lead":"This paper argues that the measured 4% branching fraction of the X(3872) into J/psi pi+ pi- actually describes a mix of a bound state and a near-threshold continuum effect, and that the bound state alone could decay to those particles far more often. It derives an upper bound of 33% for that true branching fraction and constrains a simple model of the particle's line shapes using a new inclusive production measurement from BaBar.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 33% bound in Eq. (34) assumes the measured SDD branching-ratio inputs are pure bound-state ratios; the paper neither establishes pole dominance in the experimental windows nor verifies the narrow-bound-state condition, so the bound's target is not secured.","rationale":"The paper's factorization argument is correct, and the distinction between a resonance-feature branching fraction and a bound-state branching fraction is an important, well-stated point. Eq. (34) is also a legitimate conditional bound if the input ratios really are bound-state ratios and the bound state is narrow. My concern is not that the derivation is internally inconsistent, but that the decisive premise—that the measured PDG/BESIII ratios in Eq. (33) are pure bound-state ratios rather than mixtures containing the same threshold/virtual-state contamination the paper warns about—is asserted, not demonstrated. The reader's weakest assumption (the resonance-energy definition in Eq. (14) and the arbitrary Emin/Emax window) is real but affects only the model-parameter constraints and the 2.5–3.2 ratio in Section V.C; it does not bear on the 33% upper bound. The most load-bearing unverified condition is therefore the pole isolation/narrowness behind Eq. (34). This does not overturn the verdict: the paper is already CONDITIONAL, and the condition should be sharpened to include an explicit check that the experimental ratio inputs are pole-dominated or, failing that, a conservative treatment of the continuum contamination.","tokens_in":19665,"tokens_out":19715,"duration_ms":229421,"concrete_test":"Reanalyze the published X(3872) signals in J/ψπ+π−, J/ψω, J/ψγ, ψ(2S)γ, and χc1π0 with a simultaneous fit to the amplitude in Eq. (10) convolved with the detector resolution, extracting the bound-state-pole yield separately from the continuum/threshold contribution. Recompute Eq. (33) using only pole yields. If the reciprocal central value falls below 3.03 (i.e., Br exceeds 33%) or the pole fraction in the experimental window is not close to 1, the upper bound in Eq. (34) is not established for the bound state.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is Eq. (34): Br[X→J/ψπ+π−]<33% (90% C.L.) for the bound state. It is obtained from Eq. (33), an identity among bound-state branching fractions, with inputs Br(X→J/ψω)/Br(X→J/ψπ+π−)=0.8±0.3, Br(X→J/ψγ)/Br(...)=0.24±0.05, Br(X→ψ(2S)γ)/Br(...)=2.6±0.6, and Br(X→χc1π0)/Br(...)=0.88±0.34 taken from PDG/BESIII. Those measurements are made on the X(3872) signal in B decays with finite mass windows. In the authors' own near-threshold S-wave model, every SDD line shape is proportional to |f(E)|^2 (Eq. 16), which contains a bound-state pole plus a non-negligible continuum/virtual-state contribution unless the bound state is extremely narrow. The paper does not show that the experimental windows isolate the pole; it only states the bound-state condition. Indeed §V.A warns that D0D0barπ0 product branching fractions include the threshold enhancement, and §V.B criticizes earlier 8–10% bounds as resonance-feature bounds for exactly this reason. The same objection applies to the SDD ratios unless pole dominance is demonstrated. If those inputs are feature ratios, Eq. (33) does not close on bound-state branching fractions and the 33% value is not a bound on Br[X→J/ψπ+π−]. Relatedly, the paper's own narrowness criterion Eq. (39) only requires |Re Epole|>|Im Epole|, whereas §II requires |EX| significantly larger than ΓX/2; with EX≈−0.17 MeV at 1σ below Eq. (7), the width upper bound available at the time (ΓX<1.2 MeV) does not establish that condition. The conceptual distinction between feature and bound-state branching fractions remains solid; the numerical upper bound needs an additional pole-isolation or narrowness check.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper distinguishes the branching fraction of the X(3872) \"resonance feature\" observed in B+→K+ transitions from the branching fraction of the X bound state, arguing that they need not be equal for a near-threshold S-wave resonance. After reviewing the factorization argument that a narrow bound state's branching fractions are production-independent (Sec. II), the authors introduce a single-channel model with a complex inverse scattering length (Eq. 10) and define the theoretical resonance energy as the median of the |f(E)|^2 line shape over the chosen interval [Emin, Emax] = [−7.0, +8.2] MeV (Eqs. 14–15). Using the BaBar inclusive recoil measurement, they infer a J/ψπ+π− branching fraction of (4.1 ± 1.3)% for the resonance feature (Eq. 29). Using measured ratios for four short-distance decay modes, they derive an upper bound Br[X→J/ψπ+π−] < 33% (90% C.L.) for the bound state (Eq. 34) and a corresponding 6.7% bound for the feature (Eq. 35). They then use the measured resonance energy EX = (+0.01 ± 0.18) MeV and an estimate of the constituent-decay branching fraction (Eq. 31) to constrain the model parameters, concluding that the bound-state short-distance branching fraction exceeds the feature-level one by a factor 2.5–3.2 (Sec. V.C). A Note Added argues that the contemporaneous Li–Yuan analysis [48] mixes feature-level inputs from different production mechanisms.","tokens_in":20147,"tokens_out":31058,"duration_ms":306138,"significance":"If the central distinction holds, the paper makes an important and practically relevant point: the widely quoted ~4% \"branching fraction\" of the X(3872) into J/ψπ+π− is a feature-level quantity and should not be used directly, for example, in predictions of triangle-singularity peak heights; the bound-state branching fraction is larger and is bounded by 33%. The conceptual argument in Sec. II is sound and is stated with appropriate conditions. Strengths of the manuscript include the model-independent identity Eq. (33), the transparent and easily auditable arithmetic behind the one-sided upper bounds, the analytic line-shape results in three limits (Sec. IV), the explicit statement of the assumptions (arbitrary Emin/Emax, order-of-magnitude Im[γX] ~ √(μΓ*0), the narrowness criterion), and the correct identification of the earlier 8–10% bounds in Refs. [46] and [9] as feature-level rather than bound-state bounds. The main weaknesses are that the identification of the theoretical resonance-energy prescription with the PDG value is left unquantified, the input ratios entering Eq.","major_comments":[{"comment":"The concern that the 33% bound may apply to the resonance feature rather than to the bound state lands only partially, but it is not addressed. Equation (34) is derived by inserting measured branching ratios (J/ψω, J/ψγ, ψ(2S)γ, χc1π0, each over J/ψπ+π−) into the identity Eq. (33), and its interpretation as a bound on the bound-state branching fraction requires that these measured ratios be bound-state (pole-dominated) ratios rather than feature-level ratios. The paper never states this requirement; a reader is left to wonder why the threshold-enhancement objection that the authors correctly raise against Refs. [46], [9], and [48] does not apply to their own inputs. Within the single-channel model of Eq. (10) the concern is muted, because all short-distance decay modes share the line shape |f(E)|^2 (Eq. 16), making the ratio window-independent, and the paper should say this explicitly and add the empirical observation that the X peak is the same narrow object in all four input modes. If the inputs were feature-level ratios, Eq. (34) would bound the feature branching fraction, which is smaller than the bound-state branching fraction, so the stated claim would not follow. This is fixable by adding one justification paragraph, but as written the target of the 33% bound is not secured.","section":"§V.B, Eq. (34)"},{"comment":"The quantitative constraints in Sec. V.C hinge on identifying the theoretical resonance energy, defined as the median of |f(E)|^2 over the arbitrarily chosen interval [Emin, Emax] = [−7.0, +8.2] MeV (Eqs. 14–15), with the PDG value EX = (+0.01 ± 0.18) MeV, which was extracted by a different data-fitting procedure. The systematic uncertainty from this identification is not estimated. In the bound-state limit the analytic result Eq. (22) contains corrections of order (μΓX²/(16 Re[γX]²)) times logarithmic factors; for representative parameters (Re[γX] ≈ 30 MeV, ΓX ≈ 0.5 MeV) these corrections are of order 0.1 MeV, comparable to the ±0.18 MeV experimental error. The prescription dependence also affects the feature-level branching ratios computed through Eq. (37). A sensitivity analysis varying Emin and Emax and comparing alternative definitions (line-shape maximum, Breit-Wigner fit to the peak region) should be added, and the constraints in Figs. 6–7 and the quoted factor 2.5–3.2 should be presented with the resulting uncertainty. The qualitative claim in the abstract does not depend on this identification, but the quantitative model constraints do.","section":"§IV, Eq. (14); §V.C, Figs. 6–7"},{"comment":"The paper uses different standards for \"narrow bound state\" in the general argument and in the model application. Section II defines a narrow bound state by |EX| significantly larger than ΓX/2, while the criterion adopted in the model, Eq. (39), is only |Re Epole| > |Im Epole|, which is equivalent to |EX| > ΓX/2 and admits marginal cases. The present data do not establish the stronger condition: at 1σ below the central value in Eq. (7), EX ≈ −0.17 MeV, while the experimental limit ΓX < 1.2 MeV (Ref. [47]) gives ΓX/2 of order or larger than |EX|. Within the model, Im[γX] of order √(μΓ*0) with Re[γX] ≈ 20–30 MeV, as required for a bound state with |EX| ≈ 0.2 MeV, yields ΓX/2 ≈ |EX| ≈ 0.2 MeV, i.e., a marginally narrow state even by the weak criterion. The factor 2.5–3.2 quoted in Sec. V.C and the quantitative force of the \"considerably larger\" statements rest on this marginal case; the authors should adopt the Sec. II criterion, quantify how the narrow-bound-state regions of Figs. 6–7 shrink under a stricter requirement such as |EX| > 2ΓX, or explicitly label the 2.5–3.2 ratio as an estimate valid only in the marginal-narrowness regime.","section":"§II, §V.C, Eq. (39)"}],"minor_comments":[{"comment":"The 1σ error ellipse for BF = (25 ± 40)% extends into negative values of a quantity that must be nonnegative; the ellipse, and the claim that the model curves are compatible with Im[γX]/√(μΓ*0) up to about 9, should be truncated at BF = 0.","section":"§V.A, Fig. 6"},{"comment":"In the discussion of narrow bound states, the paper should cite the experimental limit ΓX < 1.2 MeV (Ref. [47]) to make concrete that the condition |EX| ≫ ΓX/2 is not currently established by data.","section":"§II"},{"comment":"The statement that Br(X→J/ψγ)/Br(X→J/ψπ+π−) is \"determined to be 0.24 ± 0.05 by calculating the ratio of product branching fractions from B+→K+X decays [13]\" is too terse; the specific product branching fractions and the error treatment should be spelled out.","section":"§V.B"},{"comment":"The value (4.1 ± 1.3)% relies on a preliminary BaBar result from a conference presentation (Ref. [20]); the manuscript should flag that this number is preliminary, and the one-sided Gaussian interpretation of \"90% C.L.\" in Eqs. (34)–(35) should be stated explicitly.","section":"Abstract and §V.A"},{"comment":"The entry \"31.15.bt\" in the PACS list appears malformed and should be corrected.","section":"PACS numbers"},{"comment":"The phrase \"allowing the real parameter γX to have a positive imaginary part\" is confusingly worded, since γX is then no longer real; consider: \"allowing γX, which is real under exact unitarity, to acquire a positive imaginary part.\"","section":"§III, after Eq. (10)"}],"recommendation":"major_revision","confidential_remarks":"The central conceptual distinction of the manuscript is sound, and the paper is likely to be accepted after the requested changes. I do not see grounds for rejection: the issues raised are gaps in explicit justification and sensitivity analysis rather than internal inconsistencies. The authors should be encouraged to keep the conditional character of the claims (\"if X is a narrow bound state\") prominent throughout."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe one thing to know: this paper cleanly separates a distinction that keeps getting blurred. The measured ~4% J/psi pi+pi- branching fraction applies to the production-dependent resonance feature (threshold enhancement plus possible bound state), not to the X(3872) bound state itself. Braaten et al. argue that if the X is a narrow charm-meson bound state, its own J/psi pi+pi- branching fraction is larger, and they derive an upper bound of 33% (90% C.L.) from measured ratios of other decay modes. That bound is the headline number.\n\nWhat's good: the factorization argument is standard universal scattering theory, but the application here is genuinely instructive. The bound in Eq. (33)-(34) is transparent: drop the unmeasured CD term and use the measured SDD ratios. They also correctly point out that earlier 8-10% bounds from Refs. [46] and [9] are actually bounds on the resonance feature, not on the bound state. The paper is honest about the crudeness of its line-shape model and about the arbitrariness in the integration window for the theoretical resonance energy.\n\nThe soft spots: the model constraints in Sec. V come from matching a theoretical resonance energy defined as the median of |f(E)|^2 over Emin=-7.0, Emax=+8.2 MeV (a choice they admit is arbitrary) to the PDG value extracted with a different fitting procedure. Those constraints are illustrative, not solid. The input CD branching fraction has huge uncertainties, and the BaBar measurement is preliminary. The stress-test worry about pole dominance in the experimental windows is real but, within the simplest model, not fatal: all SDD modes share the same |f(E)|^2 line shape, so ratios among them are independent of how much of the feature sits inside the window. The more serious caveat is that the experimental ratios themselves come from fits that assume specific line shapes; bias from the threshold enhancement could shift those numbers. That warrants caution, not rejection.\n\nWho is this for: anyone working on X(3872) line shapes, molecular interpretations, or near-threshold resonances. I would cite it for the feature-versus-bound-state distinction. It deserves a serious referee, even though the quantitative section is the softest part. Send it to review.","headline":"A useful conceptual clarification of X(3872) branching fractions, with an honest but model-dependent upper bound that deserves a serious referee.","tokens_in":20638,"tokens_out":6296,"would_cite":true,"duration_ms":68114,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["14.80.Va","67.85.Bc","31.15.bt"],"model":"deepseek-v4-flash","headline":"If the X(3872) is a bound state of charm mesons, its branching fraction into J/ψπ+π− is considerably larger than the roughly 4% measured for the resonance feature, with an upper bound of 33%.","keywords":["X(3872)","exotic hadrons","charm mesons","charm-meson molecule","branching fraction","near-threshold resonance","line shape","triangle singularity"],"falsifier":"Measure the $B^+\\to K^+ D^0\\bar{D}^0\\pi^0$ line shape with resolution good enough to resolve a bound-state peak below the $D^{*0}\\bar{D}^0$ threshold; if the resolved bound-state peak gives a $J/\\psi\\,\\pi^+\\pi^-$ branching fraction equal to or smaller than the roughly 4% resonance-feature value, the claimed enhancement and the bound-state interpretation in this model would be falsified.","tokens_in":19437,"feed_emoji":"⚛️","tokens_out":15511,"duration_ms":140571,"temperature":0.7,"pith_summary":"This paper argues that the branching fraction of the $X(3872)$ quoted by experiments is not a property of a single particle but of a resonance feature: a near-threshold enhancement in charm-meson production that may include, but is not the same as, a narrow bound state of $D^{*0}\\bar{D}^0$ and $D^0\\bar{D}^{*0}$. If the $X$ is such a bound state, its true branching fraction into $J/\\psi\\,\\pi^+\\pi^-$ should be considerably larger than the roughly 4% measured for the whole feature, and the paper derives an upper bound of 33% at 90% confidence from measured branching ratios of other decay modes. The distinction matters because searches for triangle-singularity peaks in $X\\pi$ or $X\\gamma$ production need the bound-state branching fraction, not the feature average, to predict how tall those peaks will be.","feed_headline":"X(3872) bound-state J/ψπ+π− rate lies above 4%, below 33%","feed_subtitle":"Separating the true charm-meson-molecule decay from the resonance-feature average sharpens tests of the X(3872) nature.","key_machinery":"The argument is carried by pole factorization: near a narrow bound-state pole the scattering amplitude factorizes as $f_{ij}(E)\\approx -c_i c_j/(E-E_X+i\\Gamma_X/2)$, so production and decay amplitudes separate, and the branching fraction obtained by integrating over the narrow bound-state peak is independent of the production mechanism. The numerical constraints use the simplest plausible line-shape model, $f(E)=1/(-\\gamma_X+\\sqrt{-2\\mu(E+i\\Gamma_{*0}/2)})$, a universal near-threshold amplitude with a single complex inverse scattering length $\\gamma_X$; its real part controls whether the state is bound or virtual, and its imaginary part encodes short-distance decay modes. The paper also defines the resonance energy as the median of the $J/\\psi\\,\\pi^+\\pi^-$ line shape over the interval $E_{\\min}=-7.0$ MeV to $E_{\\max}=+8.2$ MeV and derives analytic expressions for it in the bound-state, zero-energy, and virtual-state limits.","core_discovery":"The central claim is that a near-threshold $S$-wave resonance like the $X(3872)$ has two different kinds of branching fractions that should not be confused. The resonance feature seen in $B^+\\to K^+$ transitions—a threshold enhancement plus a possible bound-state or virtual-state peak—has a $J/\\psi\\,\\pi^+\\pi^-$ branching fraction of about 4%. If the $X$ is a narrow bound state, factorization of the amplitude at the pole guarantees that its own branching fractions are production-independent, and its $J/\\psi\\,\\pi^+\\pi^-$ branching fraction is larger than the feature's. Using measured branching ratios into $J/\\psi\\omega$, $J/\\psi\\gamma$, $\\psi(2S)\\gamma$, and $\\chi_{c1}\\pi^0$, the paper puts an upper bound $\\mathrm{Br}[X\\to J/\\psi\\,\\pi^+\\pi^-]<33\\%$ at 90% confidence, and in the simplest line-shape model the bound-state branching fraction exceeds the feature's by a factor of 2.5 to 3.2 within the $1\\sigma$ region allowed by the resonance energy.","pith_inferences":["Extending the paper's logic, other near-threshold exotic candidates may also have quoted 'branching fractions' that mix a threshold enhancement with a possible bound-state peak, so those numbers should be re-examined mode by mode.","A decisive testable extension would be high-resolution D0 Dbar0 pi0 line-shape data that resolve a bound-state peak below threshold; such data could confirm or refute the predicted 2.5 to 3.2 enhancement.","I would treat the enhancement factor as an estimate from the simplest single-channel line-shape model; coupling to charged charm-meson pairs or to χc1(2P) could shift the numerical factor.","Tighter measurements of the D0 Dbar0 gamma component and of absolute B+→K+X bound-state production would turn the loose lower bound into a direct determination of the bound-state branching fraction."],"forward_implications":["The measured 4% J/ψπ+π− branching fraction of the resonance feature should not be used as the branching fraction of the X itself; if the X is a bound state, the true rate is larger, with an upper bound of 33%.","Triangle-singularity peak heights in Xπ and Xγ production scale with the bound-state branching fraction, so those peaks could be roughly 2.5 to 3.2 times higher than predictions based on the feature-average value.","Within the simplest line-shape model, the measured resonance energy and the estimated constituent-decay branching fraction constrain the inverse scattering length; most of the 1σ region corresponds to a virtual state, with narrow bound states allowed only for negative E_X.","Without the χc1π0 decay mode the upper bound would be 44%, so the measurement of X→χc1π0 is what tightens the bound to 33%.","The same distinction between a resonance feature and a bound state applies to any near-threshold S-wave resonance: feature branching fractions depend on production mechanism, while bound-state branching fractions do not."],"supporting_citations":[{"why":"supplies the measured inclusive branching fraction for B+→K+ plus the X resonance feature, the source of the 4% J/ψπ+π− value.","marker":"[20]"},{"why":"supplies the measured resonance energy and the product branching fractions for other X decay modes used in the upper-bound identity.","marker":"[13]"},{"why":"supplies a measurement of the D0 Dbar0 pi0 near-threshold enhancement used to estimate the constituent-decay branching fraction of the resonance feature.","marker":"[23]"},{"why":"introduces the simplest analytic line-shape model for the X whose amplitude f(E) is used to constrain γ_X.","marker":"[26]"},{"why":"gives the universal low-energy S-wave scattering amplitude with inverse scattering length that the line-shape model starts from.","marker":"[21]"},{"why":"provides the D*0 decay width Γ_{*0} that sets the width scale in the model and in the bound-state width formula.","marker":"[22]"},{"why":"provides the measured X→J/ψω branching ratio that enters the sum in the upper-bound identity.","marker":"[44]"},{"why":"provides the measured X→χc1π0 branching ratio that tightens the upper bound from 44% to 33%.","marker":"[45]"}],"fun_headline_variants":["X(3872) true J/ψππ decay exceeds its 4% feature rate","X(3872) bound-state rate: 4%–33% window sharpens nature test","X(3872) feature vs true decay: J/ψππ bound 4–33%","Model constrains X(3872) J/ψππ rate between 4% and 33%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative constraints assume that the theoretically defined resonance energy—the median of the $J/\\psi\\,\\pi^+\\pi^-$ line shape over the chosen interval from $-7.0$ to $+8.2$ MeV—matches the experimental $E_X=(+0.01\\pm 0.18)$ MeV, which was extracted with a different fitting procedure; if the two definitions do not correspond, the fitted model parameters and the predicted 2.5 to 3.2 enhancement are unreliable.","fun_headline_variants_meta":{"raw":{"variants":["X(3872) true J/ψππ decay exceeds its 4% feature rate","X(3872) bound-state rate: 4%–33% window sharpens nature test","X(3872) feature vs true decay: J/ψππ bound 4–33%","Model constrains X(3872) J/ψππ rate between 4% and 33%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000746,"raw_usage":{"total_tokens":3392,"prompt_tokens":1082,"completion_tokens":2310,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":698,"completion_tokens_details":{"reasoning_tokens":2205}},"tokens_in":698,"tokens_out":2310,"duration_ms":19886,"temperature":1.0,"reasoning_tokens":2205,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:34:22.574093+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the $B^+\\to K^+ D^0\\bar{D}^0\\pi^0$ line shape with resolution good enough to resolve a bound-state peak below the $D^{*0}\\bar{D}^0$ threshold; if the resolved bound-state peak gives a $J/\\psi\\,\\pi^+\\pi^-$ branching fraction equal to or smaller than the roughly 4% resonance-feature value, the claimed enhancement and the bound-state interpretation in this model would be falsified.","supporting_citations":[{"cited_title":"Isospin properties of the X state near the D {\\bar D}^{*} threshold","cited_arxiv_id":"0704.3029","evidence_quote":"introduces the simplest analytic line-shape model for the X whose amplitude f(E) is used to constrain γ_X."}],"review_version":1}