{"id":"6cb76ea0-4970-4d41-b34b-f2905a876d04","arxiv_id":"1909.01702","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A simple fluctuating-string model, fitted to one-particle rapidity spectra, quantitatively reproduces the ATLAS measurements of long-range longitudinal two-particle correlations in Pb-Pb, p-Pb, and p-p collisions.","lead":"A semi-analytic string model with fluctuating endpoints fitted to particle rapidity spectra at the LHC reproduces the two-particle longitudinal correlations measured in Pb-Pb, p-Pb, and p-p collisions. The model favors four or five wounded constituent partons per nucleon and a 'disjoint' arrangement of string endpoints.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The uniform string-breaking kernel assumed in Sec. IV is untested; because the endpoint CDFs are extracted from one-body spectra under that same kernel, a non-uniform kernel would change the predicted a11 and weaken the ATLAS comparison.","rationale":"The reader's weakest_assumption correctly identifies the uniform string-breaking kernel as the most load-bearing assumption. I agree, with a small addition: the concern is not merely that s might be non-uniform; it is that the extraction of G1/G2 from the one-body profile and the construction of the two-particle profile use the same kernel. Consequently, any alternative kernel would produce different endpoint distributions and different a11 values, so the successful comparison with ATLAS does not independently confirm the string-breaking picture; it only confirms the model within a particular kernel choice. Secondary issues noted by the reader are real but less fundamental: the p-Pb 1% centrality exclusion is post hoc, the Pb-Pb correlation data are at 2.76 TeV while the model is at 5.02 TeV, and the p-p correlation data are at 13 TeV while the same 5.02 TeV profile is used. These could be addressed by rerunning at matched energies without changing the structure of the model. The uniform-kernel issue, by contrast, affects the interpretation of every correlation prediction. A concrete computational test with a deformed kernel would settle whether the conclusion is robust. The model's simplicity and reliance on publicly available GLISSANDO-based machinery are genuine strengths, and the concern is about the specificity of the tested hypothesis rather than any internal inconsistency.","tokens_in":15211,"tokens_out":9839,"duration_ms":96629,"concrete_test":"Repeat the extraction of Secs. III-IV with a one-parameter deformed kernel, e.g. s(y;y1,y2)=omega[1+lambda(2(y-y_mid)/(y2-y1))] for y between y1 and y2, with y_mid=(y1+y2)/2 and lambda=0 for uniform, lambda=+/-0.5 for endpoint/center enhancement. Refit the same Pb-Pb and p-Pb dN/deta data at 5.02 TeV to determine the endpoint distributions for each lambda, then recompute a11(Nch) with the corresponding two-particle kernel and compare to ATLAS. If a11 changes by less than the experimental uncertainty, the uniform assumption is not load-bearing; if it shifts by more, the paper's conclusion should be softened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section IV, step 2 assumes that the string-breaking distribution s(y;y1,y2) is constant between the fluctuating endpoints y1 and y2. This assumption enters twice: Eq. (8) is used to obtain the endpoint CDFs G1 and G2 from the fitted one-body profile, and Eq. (11) uses the same uniform kernel to build the two-particle emission profile. Therefore the predicted a11 is not a parameter-free test of 'fluctuating strings' in general; it is a joint test of fluctuating endpoints together with a particular, uniform breaking law. If the true breaking rate is endpoint-enhanced or peaked at the string center, the G1/G2 values extracted under uniformity are effective objects, and the same-string contribution to the covariance in Eq. (13) changes. The comparison to ATLAS data in Figs. 10 and 12 would then not cleanly favor the disjoint endpoint scenario. The paper neither fits a shape parameter for s nor checks the stability of a11 against alternative kernels, so the central universality claim rests on this unexamined assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper extends a semi-analytic wounded-constituent string model from RHIC energies to LHC energies. The single-string emission profile f(η) is extracted point-by-point from a joint least-squares fit to Pb-Pb (ALICE) and p-Pb (ATLAS) pseudorapidity spectra at √sNN = 5.02 TeV, after generating wounded-parton numbers with GLISSANDO and overlaying a negative-binomial fluctuation. Passing from pseudorapidity to rapidity via an empirically estimated Jacobian, the authors assume uniform string breaking in rapidity and invert the one-body profile to obtain limiting endpoint distributions G1, G2 (the 'g1 = g2' and 'disjoint' cases). These are then used to compute two-particle correlations and the Legendre coefficient a11, which is compared with ATLAS data for Pb-Pb, p-Pb, and p-p collisions. The central claim, stated in Sec. VII, is that the model with 4 or 5 constituent partons per nucleon and disjoint endpoint distributions reasonably describes the data for all three systems.","tokens_in":15408,"tokens_out":4147,"duration_ms":43364,"significance":"If the central claim is correct, the paper provides a simple, almost analytic explanation of long-range longitudinal correlations at LHC energies in terms of fluctuating string endpoints, with no need for additional physics such as nuclear shadowing or baryon stopping. A genuine strength is that the two-particle correlations and a11 are not fitted to the correlation data; they are predicted from the endpoint distributions derived from the one-body spectra. The model is transparent and reproduces the order of magnitude and centrality trend of the ATLAS a11 data. However, the significance is limited by two unexamined assumptions: the uniform string-breaking kernel, which enters both the extraction of the endpoint distributions and the construction of the two-particle emission profile, and the comparison of a 5.02 TeV model with data at 2.76 TeV (Pb-Pb) and 13 TeV (p-p). The admitted large χ2 values also leave the quantitative goodness of fit unclear.","major_comments":[{"comment":"The uniform string-breaking kernel s(y; y1, y2) is used twice: Eq. (8) inverts the fitted one-body profile f(y) to obtain the endpoint CDFs G1 and G2, and Eq. (11) builds the same-string two-particle emission profile with the same kernel. The predicted a11 is therefore a joint test of fluctuating endpoints together with this particular, uniform breaking law. If the true breaking rate is non-uniform (e.g., enhanced near the endpoints or peaked at the string center), both the extracted G1, G2 and the predicted a11 change, so the ATLAS comparison in Figs. 10 and 12 would not cleanly favor the disjoint endpoint scenario. The paper neither fits a shape parameter for the kernel nor checks the stability of a11 against reasonable alternative kernels. This is a load-bearing assumption for the central universality claim and should be addressed explicitly.","section":"Sec. IV, step 2; Eqs. (8) and (11)"},{"comment":"The model parameters (single-string profile f, negative-binomial parameter q, and the number of constituent partons) are determined at √sNN = 5.02 TeV from Pb-Pb and p-Pb spectra, but the ATLAS a11 data shown in Fig. 10 for Pb-Pb are at 2.76 TeV and those in Fig. 12 for p-p are at 13 TeV. The paper states that the energy mismatch is 'numerically not significant' without providing a quantitative argument. Since the number of constituent partons itself changes with collision energy (3 at RHIC, 4–5 at LHC), the energy dependence of the model parameters is apparently not negligible. The claimed universality across systems and energies would be much strengthened by comparing with data at matched energies or by demonstrating explicitly that the relevant predictions vary only weakly over the quoted energy range.","section":"Secs. V and VI, Figs. 10 and 12"},{"comment":"The goodness of fit is not quantitatively characterized. Table I lists the least-squares values L for the 3, 4, 5, and 6 constituent variants, but no number of fitted data points, no number of parameters per point, and no normalization are given, so the reader cannot judge whether L = 140 vs 137 is a meaningful difference. The text also concedes that χ2/d.o.f. values are 'large' and 'cannot be used as stringent measures.' This makes it difficult to assess the central claim that the model 'reasonably describes' the spectra. The authors should provide a more interpretable goodness-of-fit statistic, or at least state the number of fitted points and the typical size of experimental errors.","section":"Sec. III, Table I and χ2 discussion"}],"minor_comments":[{"comment":"The exclusion of the most central 1% p-Pb data from the fit is mentioned only in a footnote; this selection should be described and justified in the main text, because it could influence the extracted f(η) and hence the endpoint distributions.","section":"Footnote 1, Sec. III"},{"comment":"The text says the Jacobian dη/dy is obtained from ALICE data for the 5% most central Pb-Pb collisions, but it is not stated whether this same Jacobian is applied to all systems (Pb-Pb, p-Pb, and p-p) and all centralities; the applicability of a single Jacobian to p-p and p-Pb should be clarified.","section":"Sec. IV, Eqs. (5)–(7)"},{"comment":"Equation (11) for the two-particle emission profile is taken from reference [1] without derivation; given that this is the main predictive object of the paper, a brief self-contained derivation or at least an intuitive explanation would improve readability.","section":"Sec. V, Eq. (11)"},{"comment":"Minor typographical issues: footnote 1 contains 'far of the optimal fit' (should be 'far from'); the text around Eq. (3) uses 'k(x;n,q)' with the negative-binomial notation and should clarify that x = 0 is removed.","section":"Miscellaneous"}],"recommendation":"major_revision","confidential_remarks":"The paper presents a simple and transparent model, and the fact that the two-particle correlations are not fitted to the correlation data is a genuine positive. The main concerns are the unexamined uniform string-breaking kernel and the energy mismatch in the comparisons; both are addressable with additional analysis. I would not reject the paper, but it needs a robustness study and a more careful quantitative assessment before the universality claim can be considered established."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, know this: the paper is a straightforward extension of the authors' own RHIC-era model to LHC, and the new result is that with 4–5 wounded constituents and disjoint string-endpoint distributions, the model's a11 coefficient comes close to ATLAS data for Pb-Pb, p-Pb, and p-p. That is a real, checkable claim, and the paper makes it without fitting the correlation data.\n\nThe good parts: the one-body emission profile is fixed from pseudorapidity spectra, then the two-particle correlations are predicted. That separation is a genuine strength. The decomposition into intrinsic string correlations versus multiplicity fluctuations (a*_11/a_11) is useful, and the preference for 4–5 constituents lines up with earlier multiplicity-fluctuation analyses. The paper is also unusually transparent: it says outright that chi^2/d.o.f. is large, that the most central 1% p-Pb bin is excluded, and that some comparisons mix collision energies.\n\nThe soft spots, in order of importance. The biggest is the uniform breaking kernel s(y;y1,y2) assumed in Sec. IV. The endpoint CDFs are extracted from the one-body profile using that kernel, and the two-particle correlation is built from the same kernel. If the form is wrong—endpoint-enhanced or centered—both the extracted endpoint distributions and the predicted a11 shift. The paper never tests this, so the agreement with ATLAS is a joint test of fluctuating endpoints plus a specific, uniform breaking law, not of the endpoint concept alone. That does not sink the paper, but the authors should have tried at least one alternative kernel shape or an explicit sensitivity argument. The energy mismatches (Pb-Pb model at 5.02 TeV versus data at 2.76 TeV, p-p at 5.02 versus 13 TeV) are disclosed, but given the model's approximate nature, a simple scaling check would be more convincing than the assertion that the mismatch is negligible. The circularity charge is softer: fitting the profile to the same systems is fine as long as the predictions are genuinely out-of-sample, which they are for the two-particle correlations.\n\nWho should read it: people working on longitudinal correlations or wounded-parton phenomenology. It is not a breakthrough, but it is a competent, useful contribution that deserves peer review rather than desk rejection. I would send it to referees, but with a request that the authors address the kernel dependence and energy sensitivity.","headline":"A clear extension of the fluctuating-string model to LHC energies, with a genuine prediction for a11, but the uniform string-breaking kernel is assumed rather than tested, so the universality claim is conditional.","tokens_in":15960,"tokens_out":3648,"would_cite":true,"duration_ms":35513,"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":"A semi-analytic string model with fluctuating endpoints reproduces the LHC data on rapidity spectra and two-particle correlations in Pb-Pb, p-Pb, and p-p collisions.","keywords":["longitudinal correlations","pseudorapidity correlations","wounded partons","strings","fluctuating endpoints","LHC","a11 Legendre coefficient","Glauber model"],"falsifier":"Measure the two-particle correlation coefficient $a_{11}$ in the same systems over a substantially wider pseudorapidity window than the current $|\\eta|<2.4$, or directly extract the shape of the two-particle emission profile $f_2$; a deviation from the model's prediction that grows with the rapidity span would indicate that the uniform string-breaking assumption is wrong.","tokens_in":14986,"feed_emoji":"📈","tokens_out":5311,"duration_ms":51359,"temperature":0.7,"pith_summary":"This paper sets out to explain the long-range longitudinal correlations measured at the LHC—the tendency for particle multiplicities in different pseudorapidity bins to move together—using a minimal mechanism: each produced hadron comes from a string stretched between two endpoint partons whose positions in rapidity fluctuate from event to event. The authors show that the one-particle rapidity spectra of Pb-Pb and p-Pb collisions fix the emission profile of these strings, and that this profile in turn delimits the possible distributions of the string endpoints. When the endpoints are assumed to occupy disjoint rapidity ranges, the predicted strength of the two-particle correlations, quantified by the Legendre coefficient $a_{11}$, matches the measured values in Pb-Pb, p-Pb, and p-p collisions. If correct, this means the dominant source of these correlations is a universal string-breaking mechanism with fluctuating endpoints, and no separate explanation such as nuclear shadowing or baryon stopping is required.","feed_headline":"String endpoints that fluctuate explain LHC particle correlations","feed_subtitle":"A model with 4–5 wounded partons per nucleon matches measured a11 coefficients in Pb-Pb, p-Pb, and p-p.","key_machinery":"The central mechanism is the longitudinally extended string with two fluctuating endpoints, one attached to each wounded constituent. For a string breaking at spatial rapidity $y$, the emission is assumed uniform between endpoints $y_1$ and $y_2$, giving the one-body profile $f(y)=\\omega\\bigl(\\tfrac12 - 2[G_1(y)-\\tfrac12][G_2(y)-\\tfrac12]\\bigr)$. The same uniform breaking law builds the two-particle emission profile $f_2(y_1,y_2)=\\omega^2\\,G_1[\\min(y_1,y_2)]\\{1-G_2[\\max(y_1,y_2)]\\} + (1\\leftrightarrow 2)$, from which the correlation function and its Legendre coefficients follow. The string endpoints are random, but the measured one-body spectra fix their cumulative distributions up to a two-parameter degeneracy; the two extreme solutions are the $g_1=g_2$ case and the disjoint case, and the data select the disjoint one.","core_discovery":"The central claim is that a semi-analytic model in which hadrons are emitted from longitudinally extended strings pulled by wounded constituents—with the string endpoints fluctuating in rapidity according to distributions $g_1$ and $g_2$—reproduces, with 4 or 5 constituent partons per nucleon, the LHC data on rapidity spectra in Pb-Pb and p-Pb at 5.02 TeV and the measured values of the correlation coefficient $a_{11}$ for Pb-Pb, p-Pb, and p-p. The one-body emission profile $f(y)$ is related to the endpoint cumulative distributions by $f(y)=\\omega\\bigl(\\tfrac12 - 2[G_1(y)-\\tfrac12][G_2(y)-\\tfrac12]\\bigr)$, so the measured spectra place bounds on the endpoint distributions; the measured correlations then select the disjoint limiting case, where the left-going and right-going endpoints live on opposite sides of the profile maximum. In that case the intrinsic string-emission correlations account for about 40% of $a_{11}$ in Pb-Pb and dominate in p-Pb.","pith_inferences":["A direct extension would be to measure $a_{11}$ over a substantially wider rapidity window; the model's specific prediction for how the coefficient scales with the rapidity span would distinguish uniform from non-uniform string breaking.","The same inversion of the one-body profile could be applied to three-particle or forward-multiplicity correlations, which would be more sensitive than $a_{11}$ to the shape of the string-breaking probability between endpoints.","The energy-dependent optimal number of constituents, if confirmed by future energy scans, would suggest that the effective partonic resolution of a nucleon grows with collision energy, a testable consequence that goes beyond the present paper's stated scope."],"forward_implications":["A single universal emission profile and endpoint distribution describe three different collision systems (Pb-Pb, p-Pb, p-p) with one set of parameters, so longitudinal correlations become a property of the string mechanism rather than of the specific nuclear size.","The extracted endpoint distributions provide definite predictions for the two-particle correlation function in pseudorapidity windows beyond the current ATLAS coverage of $|\\eta|<2.4$.","The dominance of intrinsic string correlations in p-Pb (above 80% of $a_{11}$) means that in small systems the measured $a_{11}$ is a direct probe of the string-breaking mechanism, whereas in Pb-Pb source-number fluctuations contribute nearly as much.","The absence of any need for shadowing or baryon stopping in reproducing the data suggests these effects play a minor role in the longitudinal correlations at LHC energies.","The preference for 4 or 5 constituent partons per nucleon at LHC energies, compared with the earlier 3-constituent fit at RHIC energies, indicates an energy-dependent effective number of wounded partons."],"supporting_citations":[{"why":"Supplies the earlier RHIC analysis and the method of inverting the one-body profile to constrain endpoint distributions.","marker":"[1]"},{"why":"Introduces strings with fluctuating end-points and the derivation of the one-body and two-body emission profiles used here.","marker":"[2]"},{"why":"Provides the wounded-nucleon scaling law that relates the rapidity spectrum to the number of sources from each side.","marker":"[12]"},{"why":"GLISSANDO 3 generates the wounded-parton configurations and centrality classes used in the fits.","marker":"[39]"},{"why":"The ALICE Pb-Pb pseudorapidity spectra at 5.02 TeV are one of the two data sets used in the joint fit of the emission profile.","marker":"[40]"},{"why":"The ATLAS p-Pb pseudorapidity spectra at 5.02 TeV provide the asymmetric part of the fit and the data for correlation comparison.","marker":"[41]"},{"why":"The measured ATLAS $a_{11}$ values for Pb-Pb, p-Pb, and p-p are the main quantitative comparison for the model's correlation predictions.","marker":"[45]"},{"why":"Supports the use of about 5 constituent partons per nucleon at LHC energies, matching the paper's preferred model variants.","marker":"[34]"}],"fun_headline_variants":["Fluctuating string endpoints unify LHC collision data","String model with 4-5 partons reproduces LHC correlations","String endpoint fluctuations explain Pb-Pb, p-Pb, and p-p correlations","LHC correlation coefficients matched by fluctuating string model","Unified description of LHC correlations via fluctuating string ends"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that a string breaks uniformly in rapidity between its two endpoints; if the true breaking is non-uniform, the endpoint distributions extracted from the one-particle spectra and the predicted $a_{11}$ would both change, so the comparison to data would not be a clean test of the string picture.","fun_headline_variants_meta":{"raw":{"variants":["Fluctuating string endpoints unify LHC collision data","String model with 4-5 partons reproduces LHC correlations","String endpoint fluctuations explain Pb-Pb, p-Pb, and p-p correlations","LHC correlation coefficients matched by fluctuating string model","Unified description of LHC correlations via fluctuating string ends"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001553,"raw_usage":{"total_tokens":6185,"prompt_tokens":900,"completion_tokens":5285,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":5198}},"tokens_in":516,"tokens_out":5285,"duration_ms":35975,"temperature":1.0,"reasoning_tokens":5198,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:09:27.869889+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the two-particle correlation coefficient $a_{11}$ in the same systems over a substantially wider pseudorapidity window than the current $|\\eta|<2.4$, or directly extract the shape of the two-particle emission profile $f_2$; a deviation from the model's prediction that grows with the rapidity span would indicate that the uniform string-breaking assumption is wrong.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the earlier RHIC analysis and the method of inverting the one-body profile to constrain endpoint distributions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces strings with fluctuating end-points and the derivation of the one-body and two-body emission profiles used here."},{"cited_title":"This ratio is plotted in Fig","cited_arxiv_id":null,"evidence_quote":"Provides the wounded-nucleon scaling law that relates the rapidity spectrum to the number of sources from each side."},{"cited_title":"Adare et al","cited_arxiv_id":null,"evidence_quote":"GLISSANDO 3 generates the wounded-parton configurations and centrality classes used in the fits."},{"cited_title":"Scaling properties of the mean multiplicity and pseudorapidity density in $e^{-}+e^{+}$, $e^{\\pm}$+p, p($\\bar{\\mathrm{p}}$)+p, p+A and A+A(B) collisions","cited_arxiv_id":"1601.06001","evidence_quote":"The ALICE Pb-Pb pseudorapidity spectra at 5.02 TeV are one of the two data sets used in the joint fit of the emission profile."},{"cited_title":"Bo˙ zek, W","cited_arxiv_id":null,"evidence_quote":"The ATLAS p-Pb pseudorapidity spectra at 5.02 TeV provide the asymmetric part of the fit and the data for correlation comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The measured ATLAS $a_{11}$ values for Pb-Pb, p-Pb, and p-p are the main quantitative comparison for the model's correlation predictions."},{"cited_title":"Bia las and A","cited_arxiv_id":null,"evidence_quote":"Supports the use of about 5 constituent partons per nucleon at LHC energies, matching the paper's preferred model variants."}],"review_version":1}