{"id":"62e3048a-7a95-46bb-9e5c-4d1207b78e2e","arxiv_id":"2505.02706","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In ultracentral Pb+Pb collisions, the sharp drop in event-by-event mean-pT fluctuations is attributed to vanishing impact-parameter fluctuations in a thermalized hydrodynamic medium, and v0(pT) is shown to act as a differential radial-flow observable.","lead":"When two lead nuclei smash together, the average sideways momentum of the produced particles varies from collision to collision. This paper says a sudden drop in that variation in the most head-on collisions is a fingerprint of the hot matter reaching thermal equilibrium, and it proposes the differential observable v0(pT) as a radial-flow probe.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'direct evidence of thermalization' rests on a fitted 2D Gaussian correlation r(b); the sharp variance drop is generated by that fitted correlation, so the inference is circular until r(b) is predicted rather than fit.","rationale":"I agree with the reader's weakest_assumption: the 2D correlated Gaussian with a single r(b) is the load-bearing element. The paper's own text says the fit 'returns a very large value of r' and then attributes this to thermalization; this is an inference from a fit, not a prediction. The HIJING comparison shows lack of correlation but does not establish that thermalized hydrodynamics uniquely produces the correlation, nor does it quantify how much of the drop is controlled by r versus by the multiplicity-width parameters σ_Nch(b). Because the manuscript is a short proceedings paper that sends the reader to Refs. [3,6] for details, the essential validation is absent in this text. The suggested hydro test is feasible with existing event-by-event hydro codes and directly targets the gap. I do not see a reason to reject the paper: the observable and the v0(pT) construction are interesting, and the conditional verdict is appropriate. However, the 'direct evidence' phrase should be softened unless the test is performed.","tokens_in":4084,"tokens_out":5827,"duration_ms":69004,"concrete_test":"Run event-by-event hydrodynamics at several fixed impact parameters (e.g., b=0, 2, 4, 6 fm), construct the joint distribution P([pT],Nch|b) non-parametrically, and compute Var(pT|Nch) directly from the simulated events without fitting r(b) to ATLAS data. Then check two things: (i) whether the predicted Var(pT|Nch) reproduces the steep ultracentral drop, and (ii) whether the correlation extracted from the simulated P([pT],Nch|b) matches r_Nch≈0.676. If the drop appears only after r(b) is fit to the measured variance, the thermalization inference fails; if the drop is present already in the parameter-free hydro prediction, the claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the steep decrease in Var(pT|Nch) in ultracentral collisions is caused by the rapid decline of impact-parameter fluctuations and therefore constitutes direct evidence of QGP thermalization (Conclusions). The operative decomposition is Eq. 1, which splits Var(pT|Nch) into an intrinsic term and a b-fluctuation term. In the assumed 2D correlated Gaussian model, the b-fluctuation term is generated entirely by the Pearson correlation r(b): if r(b)=0, the conditional mean ⟨[pT]|Nch,b⟩ is independent of b and the second term vanishes. The fit to ATLAS data returns r_Nch=0.676, and this large fitted correlation is then interpreted as 'stemming as a direct consequence of the thermalization of the QGP medium' (Sec. 2). The sharp drop is therefore not an independent observation: it is a consequence of the Gaussian ansatz plus the fitted r(b). The only non-thermal comparison is a qualitative HIJING scatter plot (Fig. 1 left), with no quantitative baseline and no uncertainty on r. The model details are also deferred to Refs. [3] and [6], so the present manuscript does not contain an independent derivation of the key distribution. Unless the correlated Gaussian and the fitted correlation are validated against an actual event-by-event hydrodynamic prediction at multiple b, the 'direct evidence' claim is circular rather than demonstrated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes that the steep decrease of the variance of the event-by-event mean transverse momentum per particle, [pT], observed by ATLAS in ultracentral Pb+Pb collisions is caused by the rapid suppression of impact-parameter fluctuations at fixed multiplicity, and interprets this as direct evidence of QGP thermalization. It also introduces the differential observable v0(pT), defined as the correlation between the pT-spectrum and [pT] fluctuations, and argues that v0(pT)/v0 is centrality independent, shows a mass hierarchy, and can be used to extrapolate the multiplicity-dependent variance of [pT] between pT windows. The analysis is based on a 2D correlated Gaussian model for P([pT], Nch|b) with five parameters, fitted to the ATLAS variance data, and on hydrodynamic simulations at fixed impact parameter.","tokens_in":4449,"tokens_out":9425,"duration_ms":93648,"significance":"The paper's starting point, Eq. (1), is an exact law-of-total-variance decomposition, and the model reproduces the shape of the ATLAS variance data. If the thermalization interpretation is correct, the observable provides a momentum-magnitude-based signature of QGP thermalization complementary to azimuthal anisotropic flow, and v0(pT) is a potentially useful differential radial-flow observable. A strength is the explicit comparison with a non-thermal HIJING baseline and the use of hydrodynamic simulations. However, the central claim hinges on the fitted correlation coefficient r(b), and the current manuscript does not establish that r(b) is predicted rather than adjusted, nor does it provide uncertainties or goodness-of-fit measures; the HIJING comparison is only qualitative. The v0(pT) extrapolation in Fig. 2 is presented without quantitative errors.","major_comments":[{"comment":"The central conclusion that the sharp variance drop is caused by impact-parameter fluctuations and constitutes direct evidence of QGP thermalization rests on the 2D correlated Gaussian ansatz for P([pT], Nch|b), with the Pearson correlation r(b) treated as a free parameter. In this model, the b-fluctuation term in Eq. (1) is equal to Var_b( r(b) [sigma_pT(b)/sigma_Nch(b)] [Nch - Nch(b)] | Nch ), so it vanishes identically if r(b)=0. The fitted value r_Nch=0.676 is adjusted to reproduce the same ATLAS data that are then interpreted as evidence for thermalization, making the inference circular unless r(b) is independently predicted. I request a validation in which r(b) (or the full conditional distribution) is obtained from event-by-event hydrodynamic simulations at several b values without fitting to the variance data, together with a quantitative goodness-of-fit estimate (e.g., chi^2 per degree of freedom) and uncertainties on the fitted parameters.","section":"Section 2, Eq. (1), Fig. 1"},{"comment":"The comparison with the non-thermal HIJING model is only qualitative. The paper states that HIJING lacks the strong correlation between [pT] and Nch, but it reports no quantitative correlation coefficient for HIJING, no uncertainty, and no resulting prediction for Var(pT|Nch) from HIJING. Without these numbers, the claim that the large r is missing in the non-thermal model and therefore signals thermalization is not established. Please provide the Pearson correlation for HIJING at fixed b and the HIJING-based variance prediction overlaid on Fig. 1 (right).","section":"Section 2, Fig. 1 (left)"},{"comment":"The reproduction of the ATLAS c_k (variance) data for the lower pT windows using the acceptance correction factor C_A is not quantified. No residuals, chi^2, or uncertainties are reported, so it is unclear how accurately the model reproduces the data. Please provide a quantitative comparison and specify how C_A is computed from v0(pT)/v0 for each pT window, including the propagation of uncertainties from the fitted parameters.","section":"Section 3, Fig. 2 (right)"},{"comment":"The model parameters (Nch(b), sigma_Nch(b), pT(b), sigma_pT(b), r(b)) are not fully specified in the text; in particular, the b-dependence of each parameter and whether r is assumed constant (as the single quoted r_Nch suggests) are not stated. Since the conclusion depends on the b-dependence of these parameters, please present the explicit parameterizations used in the fit, or refer precisely to the equations in Ref. [3] where they are defined.","section":"Section 2"}],"minor_comments":[{"comment":"Typo: 'visocous' should be 'viscous' in the first sentence.","section":"Introduction"},{"comment":"The label 'pT(0) = 9.357 MeV/c' is unphysical for the mean [pT]; if this value is actually sigma_pT(0), please relabel it.","section":"Fig. 1"},{"comment":"The legend does not explicitly specify which curve corresponds to 'Intrinsic' and which to 'b fluctuation'; please add clear labels.","section":"Fig. 1 (right)"},{"comment":"The definition of v0(pT) uses delta N(pT) and delta pT without specifying whether averages are over all events or events in a fixed centrality/multiplicity bin; please define the averaging convention.","section":"Eq. (2)"},{"comment":"The statement that v0(pT)/v0 is independent of centrality (or b) is supported by only a few simulation points; please quantify the variation or give a numerical range.","section":"Section 3, Fig. 2 (left)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is very concise and appears to be a proceedings-style contribution; the Gaussian model and v0(pT) are developed in detail in Refs. [3] and [6]. For a journal publication, the thermalization claim requires the additional validation requested in the major comments, especially the independent determination of r(b). The fit parameters are reported without uncertainties, and the HIJING baseline is not quantitative. I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this as a conference-proceedings condensation of the authors' own prior papers plus an interpretive claim. The most honest summary is: Eq. 1 is an exact law-of-total-variance decomposition; the model reproduces the ATLAS variance shape; the v0(pT) construction gives a differential radial-flow-like observable with a real consistency check; but the \"direct evidence of thermalization\" language goes beyond what the analysis establishes.\n\nThe v0(pT)/v0 acceptance-correction result is the part I trust most. Using one pT window as input and reproducing the variance in two other windows is a nontrivial, falsifiable check that does not depend on the thermalization interpretation. The left panel of Fig. 2 also shows useful insensitivity to viscosity details. That deserves credit.\n\nThe soft spot is exactly where the stress-test note points. In the correlated-Gaussian model, the impact-parameter term in Eq. 1 is generated by the Pearson correlation r(b). The sharp ultracentral drop therefore comes from the fitted value r_Nch = 0.676 combined with the Gaussian ansatz, fitted to the same ATLAS data. Calling that \"direct evidence of thermalization\" is circular unless r(b) is shown to emerge from an event-by-event hydrodynamic prediction rather than being fit. No uncertainties on r or the other parameters are reported, no goodness-of-fit is given, and the HIJING baseline appears only as a qualitative scatter plot. These omissions matter because the whole interpretive weight sits on r.\n\nI do not think this is a fatal flaw. The thermalization interpretation is plausible and consistent with the hydro picture; the problem is that the paper does not yet distinguish a fitted description from a demonstrated mechanism. The heavy reliance on refs. [3] and [6] for the model details is fine for a proceedings note, but it makes the genuinely new content here thin. The citation pattern itself is appropriate: self-citations point to the papers where the model was actually derived.\n\nWho gets value from this? Someone working on event-by-event fluctuations in heavy-ion collisions who wants a quick map of the argument and a reminder of the v0(pT) formalism. It is not a complete research paper, and the central claim should be treated as conditional until the authors provide parameter uncertainties, a quantitative non-thermal baseline, and preferably a multi-b hydrodynamic validation of the Gaussian ansatz.\n\nMy recommendation: send it to peer review rather than desk-reject, but require those additions if it is to carry the \"direct evidence\" claim. The v0(pT) piece alone merits referee time, and the fluctuation-vs-thermalization question is worth settling properly.","headline":"A compact proceedings summary of earlier work, with a plausible but oversold thermalization interpretation: the sharp variance drop is produced by a fitted correlation, not an independent prediction.","tokens_in":4948,"tokens_out":1355,"would_cite":false,"duration_ms":19047,"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":"Steep drop in transverse momentum fluctuation variance signals quark-gluon plasma thermalization.","keywords":["heavy-ion collision","transverse momentum fluctuations","quark-gluon plasma","thermalization","collective dynamics","radial flow","v0(pT)","ultracentral collisions"],"falsifier":"Take the same ultracentral Pb+Pb events and compute the third and fourth cumulants of [pT] at fixed Nch; the Gaussian ansatz predicts they are negligible, so sizable values would mean the variance decomposition in Eq. (1) is contaminated and the sharp drop could have a different origin. Equally decisive, running a nonthermal event generator with the same multiplicity selection should not produce the steep ultracentral variance fall if the paper's thermalization mechanism is correct.","tokens_in":3905,"feed_emoji":"🌀","tokens_out":5706,"duration_ms":68169,"temperature":0.7,"pith_summary":"This paper claims that event-by-event fluctuations of the mean transverse momentum per particle ([pT]) in ultracentral lead-lead collisions carry a direct, momentum-based signature of quark-gluon plasma thermalization. In the ultracentral limit the variance of [pT] at fixed charged-particle multiplicity drops sharply, and the paper attributes that drop to the rapid disappearance of impact-parameter fluctuations, which is only expected if the produced medium has thermalized and evolved hydrodynamically. The paper also introduces v0(pT), the correlation between the pT spectrum and [pT], as a differential map of these fluctuations, and argues that v0(pT)/v0 behaves like a true radial-flow observable, in the same way anisotropic flow probes collective dynamics. A sympathetic reader should care because the observable depends only on particle momenta, not azimuthal directions, giving an independent handle on QGP formation.","feed_headline":"Steep momentum variance drop marks quark-gluon plasma thermalization","feed_subtitle":"A sharp fall in event-by-event momentum fluctuation variance traces vanishing impact-parameter scatter, revealing a thermalized medium.","key_machinery":"The engine is a five-parameter two-dimensional correlated Gaussian distribution P([pT], Nch|b) for the joint event-by-event behavior of [pT] and charged multiplicity at fixed impact parameter, together with the variance decomposition in Eq. (1): Var(pT|Nch) equals the intrinsic variance at fixed b plus the variance of the mean [pT] over the impact-parameter distribution at fixed Nch. This decomposition turns the data curve into two physically separated contributions and isolates the b-fluctuation part that collapses in ultracentral collisions. The Pearson correlation r(b) between [pT] and Nch at fixed b is the load-bearing parameter that distinguishes a thermalized hydrodynamic medium from a nonthermal one. The differential observable v0(pT) is defined as the correlation of the normalized pT spectrum with [pT] fluctuations, normalized by N0(pT)σpT, and its scaled form v0(pT)/v0 is centrality independent.","core_discovery":"The central claim is that the steep decrease of Var([pT]) with multiplicity in ultracentral Pb+Pb collisions is caused by the rapid decline of impact-parameter fluctuations at fixed multiplicity, and this decline is a natural consequence of the thermalization of the QGP medium. In the paper's model, the variance at fixed Nch has two parts: an intrinsic fluctuation at fixed impact parameter and a contribution from impact-parameter fluctuations at fixed Nch; in ultracentral events the second term vanishes, producing the observed sharp fall. A large Pearson correlation r(b) between [pT] and Nch at fixed b, present in hydrodynamic simulations and absent in a nonthermal model, is identified as the microscopic origin of the effect. The companion observable v0(pT), built from the correlation between normalized spectra and [pT] fluctuations, is shown to be centrality-independent when scaled by v0, to have flow-like mass ordering for identified particles, and to enable acceptance corrections that reproduce measured variances in different pT windows.","pith_inferences":["A sharper test than the variance shape would be to measure r(b) directly by binning events in narrow Nch intervals at fixed centrality and correlating [pT] with Nch; the paper's mechanism implies this correlation should be large in ultracentral data and near zero in nonthermal models.","The Gaussian ansatz for P([pT], Nch|b) predicts that higher cumulants of [pT] at fixed Nch are negligible; measuring skewness or kurtosis in existing data would either confirm the ansatz or show where the thermalization attribution needs refinement.","If v0(pT) is truly radial flow, identified-particle v0(pT) should show a mass hierarchy that shifts with centrality in the same way v2 does, a check that could be done with already published spectra.","The mechanism also suggests that in smaller collision systems the ultracentral variance fall should be weaker or absent, providing a qualitative discriminator for whether a tiny droplet thermalizes."],"forward_implications":["If the central claim is right, the ultracentral fall in Var([pT]) is a direct, momentum-based thermalization probe that does not rely on azimuthal anisotropies.","The scaled observable v0(pT)/v0 can serve as a centrality-independent differential radial-flow observable, with mass ordering for identified particles similar to elliptic flow.","Acceptance corrections derived from v0(pT)/v0 allow prediction of σpT in arbitrary pT windows from one reference measurement; the paper reproduces measured variances in two other pT windows this way.","Shear viscosity has negligible effect on v0(pT)/v0, while bulk viscosity produces small effects at high pT when the observable is plotted against pT/⟨pT⟩.","The same logic can be extended to small systems such as p+Pb and p+p as a test of whether QGP-like collectivity and thermalization appear there."],"supporting_citations":[{"why":"Supplies the measured variance data and the five-parameter correlated-Gaussian model fit used to extract r(b) and decompose the variance.","marker":"[3]"},{"why":"Establishes the impact-parameter fluctuation contribution at fixed multiplicity, the second term in Eq. (1).","marker":"[4]"},{"why":"Proposes the correlation between particle spectra and [pT] from which v0(pT) is constructed.","marker":"[5]"},{"why":"Establishes v0(pT)/v0 as a centrality-independent radial-flow observable and provides the acceptance-correction formalism.","marker":"[6]"},{"why":"Supplies the acceptance correction factor CA used to reproduce measured variances in different pT windows.","marker":"[10]"}],"fun_headline_variants":["Momentum variance plunge reveals QGP thermalization","Thermalization written in steep [pT] variance drop","v0(pT) probes QGP flow through momentum fluctuations","Sharp [pT] variance fall maps quark-gluon plasma thermalization","Steep variance drop in ultracentral collisions signals QGP thermalization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion collapses if the joint distribution of [pT] and Nch at fixed impact parameter is not a two-dimensional correlated Gaussian with a single correlation coefficient r(b), because every extraction of the impact-parameter contribution and the thermalization signature passes through that ansatz.","fun_headline_variants_meta":{"raw":{"variants":["Momentum variance plunge reveals QGP thermalization","Thermalization written in steep [pT] variance drop","v0(pT) probes QGP flow through momentum fluctuations","Sharp [pT] variance fall maps quark-gluon plasma thermalization","Steep variance drop in ultracentral collisions signals QGP thermalization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000153,"raw_usage":{"total_tokens":1152,"prompt_tokens":836,"completion_tokens":316,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":452,"completion_tokens_details":{"reasoning_tokens":230}},"tokens_in":452,"tokens_out":316,"duration_ms":3925,"temperature":1.0,"reasoning_tokens":230,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:42:53.949781+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the same ultracentral Pb+Pb events and compute the third and fourth cumulants of [pT] at fixed Nch; the Gaussian ansatz predicts they are negligible, so sizable values would mean the variance decomposition in Eq. (1) is contaminated and the sharp drop could have a different origin. Equally decisive, running a nonthermal event generator with the same multiplicity selection should not produce the steep ultracentral variance fall if the paper's thermalization mechanism is correct.","supporting_citations":[],"review_version":1}