{"id":"368f2f11-075f-440d-9de4-b6b555abb065","arxiv_id":"2505.12961","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The thesis uses event-by-event fluctuations and correlations of collective flow and mean transverse momentum to propose new probes of the initial state, explain the ATLAS [pT]-variance fall, and constrain nuclear deformation.","lead":"This doctoral thesis studies the hottest droplet of fluid made in heavy-ion collisions by tracking event-by-event fluctuations and correlations of flow and transverse momentum. It proposes new correlation observables, explains the ATLAS variance drop in ultracentral Pb+Pb, and predicts skewness and kurtosis that experiments can test.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The steep-fall explanation hinges on a two-dimensional Gaussian ansatz whose parameters are fit to the variance; the claimed robust skewness/kurtosis predictions are untested extrapolations and should be verified against data.","rationale":"The paper compiles six peer-reviewed papers; the factorization-breaking formalism is a useful observable and the hydro comparisons are reasonable. However, the headline explanation of the ATLAS steep fall is a fit, and the model's claim to yield robust predictions for skewness and kurtosis goes beyond what the variance fit can determine. The reader's conditional verdict is appropriate: the central claim should be reframed around parameter dependence, and the higher-order predictions should be checked against data before being called robust. Agreement with the reader's weakest assumption is full; no additional fatal flaw was identified, and the published core is not internally inconsistent. Hence the verdict remains CONDITIONAL (UNCHANGED).","tokens_in":56138,"tokens_out":7121,"duration_ms":76411,"concrete_test":"Measure the skewness and kurtosis of the [pT] distribution in ultracentral (0-1%) Pb+Pb collisions from ATLAS or ALICE data in the same centrality and pT-acceptance windows used in Sec. 4.1.5, and compare with the Sec. 4.2.3 predictions. If the measured third and fourth standardized cumulants are consistent with zero within uncertainties, the non-Gaussian predictions are refuted while the variance fit may still look acceptable; if they match the predicted sign and magnitude, the impact-parameter mechanism is supported. This single comparison separates the fitted variance explanation from the claimed robust higher-order predictions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing assumption is that the joint event-by-event distribution of charged multiplicity Nch and [pT] is a bivariate Gaussian whose parameters are inferred from the ATLAS variance (Sec. 4.1.3-4.1.4). For fixed parameters, the conditional variance Var([pT]|Nch) of a bivariate Gaussian is independent of Nch; the observed steep fall therefore has to be generated by letting the parameters, especially the correlation and the impact-parameter width, vary with Nch. The fit then has enough freedom to reproduce the variance fall without independently constraining the mechanism. The skewness and kurtosis predictions (Sec. 4.2) are advertised as robust, but a Gaussian has zero skewness and kurtosis 3; their non-Gaussian content is entirely an artifact of the assumed distribution of impact parameters and the assumed linear relation between Nch and [pT]. The variance data do not constrain the tails, so these predictions are not robust unless validated by data or by a full model with the same parameters. If the true distribution has non-Gaussian tails or the correlation drifts across the ultracentral centrality range, the variance fit can remain good while the skewness/kurtosis predictions fail.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This doctoral thesis studies event-by-event fluctuations and correlations of collective observables in ultrarelativistic heavy-ion collisions, focusing on mean transverse momentum per particle [pT] and harmonic flow coefficients v_n. The main results are organized in three chapters: (i) factorization-breaking coefficients between flow vectors in different p_T bins, including experimentally feasible versions with one momentum-averaged flow, and a decomposition into flow-magnitude and flow-angle decorrelation; (ii) an explanation of the steep fall of Var([pT]) in ultracentral Pb+Pb collisions reported by ATLAS, based on a correlated two-dimensional Gaussian model for (N_ch, [pT]) with impact-parameter fluctuations, together with predicted skewness and kurtosis; and (iii) Pearson correlation coefficients between [pT] and v_n^2, higher-order normalized and symmetric cumulants, momentum-dependent correlations sensitive to nucleon width, and applications to nuclear deformation. The results are obtained with standard Glauber or TRENTo initial conditions followed by 2+1D viscous hydrodynamic evolution with MUSIC, and are compared with ALICE and ATLAS data.","tokens_in":56349,"tokens_out":3436,"duration_ms":37866,"significance":"If the results hold, the thesis makes several useful contributions: the factorization-breaking coefficients in Sec. 3.2 provide new, experimentally accessible observables that separate magnitude and angle decorrelation; the skewness and kurtosis predictions in Sec. 4.2 are falsifiable and would constrain the initial-state mechanism behind [pT] fluctuations; and the correlation/cumulant constructions in Chapter 5 give additional handles on initial-state granularity and nuclear deformation. The thesis is based on a series of published papers, and it reproduces the relevant published figures with the hydrodynamic setups. The use of explicit model-to-data comparisons for the factorization-breaking coefficients, and the analytic derivations for the correlated Gaussian variance in Chapter 4, are strengths. However, the central explanatory claim of Chapter 4 is weakened by the fact that the Gaussian parameters are fitted to the very ATLAS variance the model is said to explain, and the advertised skewness/kurtosis predictions are extrapolations of that fitted ansatz.","major_comments":[{"comment":"The abstract's central claim that the model \"can explain the steep fall\" of Var([pT]) is not supported at the level of an explanation, because the parameters of the two-dimensional Gaussian (multiplicity width, [pT] width, and correlation) are fitted to the ATLAS variance data in Sec. 4.1.5. Agreement with the fitted variance is therefore partly by construction. Moreover, for a bivariate Gaussian, Var([pT]|N_ch) is independent of N_ch, so the observed centrality/ultracentral fall must be generated by allowing the parameters to vary with N_ch; the fit has enough freedom to absorb the effect without independently constraining the physical mechanism. Please reframe the claim as a successful two-parameter description and provide an independent cross-check, for example by fixing the Gaussian parameters from separate moments or from a different centrality range before comparing with the steep fall.","section":"Sec. 4.1.3–4.1.5"},{"comment":"The skewness and kurtosis predictions are advertised as \"robust\" (abstract and Sec. 4.2.3), but a bivariate Gaussian has zero skewness and kurtosis equal to 3; all non-Gaussian content in the results comes from the assumed distribution of impact parameters and the assumed linear relation between N_ch and [pT]. The variance fit does not constrain the tails of these distributions, so the skewness and kurtosis predictions are untested extrapolations. To make the claim robust, the authors should test sensitivity to alternative impact-parameter distributions or centrality-selection prescriptions, or compare directly with experimental data; otherwise the predictions should be presented as model-dependent estimates rather than robust predictions.","section":"Sec. 4.2.1–4.2.3"},{"comment":"The statement that flow-vector decorrelation \"can be attributed to equal contributions from the flow magnitude and flow angle decorrelation\" is formulated as a general result, but the evidence is an approximate equality observed in the specific Glauber+MUSIC and TRENTo+MUSIC calculations and in a toy model. The thesis itself notes that non-flow correlations can modify the comparison with data, and the decomposition into equal halves may depend on the harmonic, centrality, and the v_n^4 weighting used in the angle correlator. Please qualify this as a model-level approximate relation rather than a universal property, or provide a derivation that states the conditions under which the relation holds.","section":"Sec. 3.2.2, Eq. (3.61) and (3.68)"}],"minor_comments":[{"comment":"The caption refers to \"Fig. 3.59\" where the intended cross-reference appears to be a later figure in the same section; this should be corrected.","section":"Fig. 3.13 caption"},{"comment":"The TRENTo model is spelled inconsistently as \"TRENTo\" and \"TRENTO\"; please unify the notation. Also, the symbol p is used both for transverse momentum and for the TRENTo reduced-thickness parameter, which is confusing in places; a distinct symbol for the model parameter would improve readability.","section":"Throughout"},{"comment":"The sentence introducing the TRENTo acronym is garbled: \"TRENTo which reads asReduced Thickness Event-by-event Nuclear Topology\" is missing punctuation and a space. Please revise the sentence.","section":"Sec. 2.6.1"},{"comment":"The comparison between the factorization-breaking coefficient and the Pearson correlation coefficient would be clearer if the figure legend explicitly identified which curve corresponds to Eq. (3.62) and which to Eq. (3.60); the current caption requires the reader to infer this from the text.","section":"Eq. (3.62) and Fig. 3.16"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a compilation of previously published work, and the new framing in Chapter 4—especially the word \"explain\" in the abstract—needs to be reconciled with the fitted nature of the Gaussian model. The remaining chapters are largely sound and the model-data comparisons are useful, but the load-bearing explanatory claim of Chapter 4 and the robustness claim for the skewness/kurtosis predictions require substantial revision. The editor may also wish to consider whether the journal's policy requires a clearer statement that most results have appeared in earlier articles by the author and collaborators."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a doctoral thesis assembled from six peer-reviewed papers. If you want the unified story and the definitions of the new correlation observables, it is a fine place to look. But the abstract's claim that the model 'can explain the steep fall' of the ATLAS variance is weaker than it sounds: the correlated-Gaussian parameters are fitted to that same variance in Sec. 4.1.5, so the 'explanation' is not a parameter-free prediction. The skewness and kurtosis predictions in Sec. 4.2 are extrapolations of that fitted Gaussian; a Gaussian has zero skewness and fixed kurtosis, so their non-Gaussian content comes entirely from the assumed impact-parameter distribution, which the variance data do not constrain. That is a real soft spot, and the stress-test note is on target.\n\nWhat the thesis does well is organize a set of useful observables: factorization-breaking coefficients between flow vectors in different pT bins (including the momentum-averaged version that eases experimental measurement), Pearson correlations between [pT] and v_n^2, higher-order symmetric cumulants, and the deformation studies in U+U. These come from the author's published papers (PRC 104-109), and the hydro simulations are standard, with comparisons to ALICE and ATLAS data. The momentum-dependent correlation's sensitivity to the nucleon width is a concrete, useful result.\n\nThe main soft spot, beyond Chapter 4, is that the thesis is explicitly a compilation: it does not present new unpublished results. That is fine for a dissertation, but for an arXiv item it limits the novelty. Also, no code or parameter files are provided, so the quantitative claims are not independently reproducible from the manuscript itself. The 'thermalization at the femtoscale' wording in the associated publication overreaches what a Gaussian fit can tell you; the thesis repeats that framing.\n\nFor a referee: the underlying science is mostly sound, and the Chapter 4 issue is a legitimate request for reframing rather than a fatal flaw. I would send this to peer review if it were submitted to a journal, with the expectation that Chapter 4 be rewritten to emphasize the fitted-parameter dependence and that the skewness/kurtosis predictions be explicitly flagged as untested until data check them. As it stands, the thesis is a reasonable resource for someone entering the field or wanting the definitions in one place.","headline":"A competent thesis compilation of six published papers; the headline ATLAS-variance explanation is fit-based, and the skewness/kurtosis predictions are untested extrapolations.","tokens_in":56929,"tokens_out":2410,"would_cite":false,"duration_ms":24148,"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 thesis argues that event-by-event fluctuations of mean transverse momentum and harmonic flow in ultracentral Pb+Pb collisions are captured by a correlated two-variable Gaussian model, and that flow-vector decorrelation between…","keywords":["quark-gluon plasma","mean transverse momentum fluctuations","harmonic flow","factorization breaking","flow decorrelation","ultracentral heavy-ion collisions","nuclear deformation","initial-state fluctuations"],"falsifier":"Measure the skewness of the [pT] distribution in ultracentral Pb+Pb events; if its sign or centrality dependence disagrees with the correlated-Gaussian prediction presented in Chapter 4, the central model is ruled out even though the variance fit may look good. A second, independent check is to bin by a centrality estimator that does not rely on charged multiplicity and test the predicted impact-parameter-fluctuation contribution to Var([pT]).","tokens_in":55862,"feed_emoji":"🔥","tokens_out":9032,"duration_ms":88302,"temperature":0.7,"pith_summary":"This thesis argues that event-by-event fluctuations of the hot, dense fluid formed in heavy-ion collisions can be read off from correlations among a handful of bulk observables: the per-event mean transverse momentum [pT], the harmonic flow coefficients v_n, and the charged multiplicity Nch. Its central demonstration is that the surprisingly steep drop in the variance of [pT] measured by ATLAS in ultracentral Pb+Pb collisions is naturally produced by a model in which Nch and [pT] fluctuate together as a two-dimensional Gaussian, with most of the ultracentral signal coming from impact-parameter fluctuations. The same framework yields skewness and kurtosis predictions for [pT]. For flow, it shows that event-by-event fluctuations break the factorization of flow vectors between different transverse-momentum bins, and that this breaking splits roughly equally into flow-magnitude and flow-angle decorrelation—a pattern that a toy model explains. If these claims hold, the observables put new constraints on the initial state and on nuclear deformation parameters.","feed_headline":"Two-variable Gaussian model explains ultracentral [pT] variance drop","feed_subtitle":"The correlated Gaussian reproduces ATLAS data and yields testable skewness, kurtosis, and flow-decorrelation predictions.","key_machinery":"The main object of the thesis is a family of correlation coefficients built from per-event flow vectors. The first-order factorization-breaking coefficient $r_n(p_1,p_2)=\\langle V_n(p_1)V_n^*(p_2)\\rangle/\\sqrt{\\langle v_n^2(p_1)\\rangle\\langle v_n^2(p_2)\\rangle}$ measures how much flow vectors in two $p_T$ bins decorrelate; a momentum-averaged version $r_n(p)=\\langle V_n V_n^*(p)\\rangle/\\sqrt{\\langle v_n^2\\rangle\\langle v_n^2(p)\\rangle}$ removes the statistics problem of needing two particles in the same bin. Second-order analogues built from $V_n^2$ allow the decorrelation to be split into magnitude and angle parts. For $[p_T]$, the workhorse is a correlated two-dimensional Gaussian $P(N_{\\rm ch},[p_T])$ whose parameters are fitted to the ATLAS variance; from it the conditional variance ${\\rm Var}([p_T]\\mid N_{\\rm ch})$ is derived, and skewness and kurtosis of $[p_T]$ are predicted. This combination of flow-vector correlators and a Gaussian moment model carries the whole argument.","core_discovery":"On the paper's own terms, the central claim is that event-by-event fluctuations of [pT] and v_n in ultracentral Pb+Pb collisions are not noise but information: they encode the same initial-state fluctuations that set the size and shape of the fireball. In the model developed here, the joint distribution of charged multiplicity Nch and [pT] is a two-dimensional Gaussian whose covariance is fixed using the ATLAS variance data, and the steep fall of Var([pT]) with centrality is then a direct consequence, with impact-parameter fluctuations playing the dominant role. The thesis further claims that fluctuations of harmonic flow can be probed by factorization-breaking coefficients between flow vectors in different pT bins, and that the resulting decorrelation decomposes into roughly equal flow-magnitude and flow-angle decorrelation. Correlations of [pT] with $v_n^{2}$, including higher-order symmetric cumulants, are presented as maps of initial-state shape-size correlations, and the same observables are shown to constrain nuclear deformation in U+U collisions.","pith_inferences":["A direct extension would be to measure the conditional variance Var([pT]|Nch) at fixed charged multiplicity; the correlated-Gaussian model makes a precise prediction for this conditional moment, and it is not the same as the centrality-binned variance shown by ATLAS, so it would be an independent test.","If the same two-dimensional Gaussian construction is applied to smaller systems such as p+Pb or low-multiplicity Pb+Pb, it would likely break down; observing where it breaks could quantify how non-Gaussian the initial-state fluctuations become.","If the near-equal split between magnitude and angle decorrelation persists event-class by event-class, experimental analyses could save statistics by measuring only the easier magnitude correlator and reconstructing the vector decorrelation from it; if it fails, that itself signals correlations between flow magnitude and angle that the toy model averages away."],"forward_implications":["The steep fall of Var([pT]) in ultracentral Pb+Pb becomes a natural consequence of impact-parameter fluctuations in a correlated Gaussian model, so no exotic source of [pT] fluctuation is needed.","The predicted skewness and kurtosis of [pT] can be checked against future measurements and would give independent constraints on the initial-state size-shape correlation.","The factorization-breaking coefficients r_n(p), r_{n;2}(p), and F_n(p) provide experimental probes of flow fluctuations, with the near-equal split between magnitude and angle decorrelation as a testable signature.","Pearson correlations between [pT] and v_n^2, together with higher-order symmetric cumulants, map initial-state shape-size correlations and add constraints beyond the usual flow measurements.","The same correlation and fluctuation observables, applied to U+U collisions, can place robust constraints on nuclear deformation parameters."],"supporting_citations":[{"why":"Defines the factorization-breaking coefficient r_n(p1,p2) that probes event-by-event flow fluctuations between transverse-momentum bins.","marker":"[240]"},{"why":"Supplies the construction of second-order and mixed-flow factorization-breaking coefficients and the main flow-decorrelation results.","marker":"[125]"},{"why":"Supplies the higher-order cumulants of [pT] and harmonic flow that underpin the correlation analyses.","marker":"[126]"},{"why":"Supplies the non-Gaussian impact-parameter-fluctuation model that explains the steep fall of Var([pT]) and predicts skewness and kurtosis.","marker":"[127]"},{"why":"Supplies the fit of the correlated-Gaussian model to the ATLAS variance data and the thermalization argument.","marker":"[128]"},{"why":"Supplies the momentum-dependent [pT]-flow Pearson correlator and its sensitivity to the nucleon Gaussian width.","marker":"[129]"},{"why":"Supplies the application of these correlations to deformed U+U collisions and constraints on nuclear deformation.","marker":"[130]"},{"why":"Provides the toy model (Appendix A.1) from which the equal magnitude/angle split of flow-vector decorrelation follows.","marker":"[246]"},{"why":"Provides the ALICE decorrelation data used as the experimental comparison for r_n(p) and flow-angle decorrelation.","marker":"[87]"}],"fun_headline_variants":["Fluctuations expose size and shape of the hottest fluid droplet","Gaussian model decodes event-by-event noise in heavy-ion data","Flow decorrelation splits into magnitude and angle contributions","Ultracentral [pT] variance fall traced to impact parameter jitter","Nuclear shape and size read from collective variable correlations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result collapses if the joint event-by-event distribution of Nch and [pT] is not close to a two-dimensional Gaussian: the model's variance explanation and its skewness and kurtosis predictions both come from that Gaussian, and its parameters are fitted to a single experimental variance curve.","fun_headline_variants_meta":{"raw":{"variants":["Fluctuations expose size and shape of the hottest fluid droplet","Gaussian model decodes event-by-event noise in heavy-ion data","Flow decorrelation splits into magnitude and angle contributions","Ultracentral [pT] variance fall traced to impact parameter jitter","Nuclear shape and size read from collective variable correlations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000518,"raw_usage":{"total_tokens":2568,"prompt_tokens":1063,"completion_tokens":1505,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":679,"completion_tokens_details":{"reasoning_tokens":1422}},"tokens_in":679,"tokens_out":1505,"duration_ms":12463,"temperature":1.0,"reasoning_tokens":1422,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:23:28.695825+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the skewness of the [pT] distribution in ultracentral Pb+Pb events; if its sign or centrality dependence disagrees with the correlated-Gaussian prediction presented in Chapter 4, the central model is ruled out even though the variance fit may look good. A second, independent check is to bin by a centrality estimator that does not rely on charged multiplicity and test the predicted impact-parameter-fluctuation contribution to Var([pT]).","supporting_citations":[],"review_version":1}