{"id":"50096ed4-df13-49ab-bbe7-1e4a194bb304","arxiv_id":"2505.15161","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In 13 TeV proton-proton collisions, the event-by-event mean transverse momentum correlation strength falls with multiplicity in a power-law fashion, with the index depending on the selected pT window.","lead":"This ALICE analysis measures how event-by-event average particle momentum fluctuates in proton-proton collisions at 13 TeV, over wider momentum windows than previous studies. It reports a power-law decrease with multiplicity and compares the trend with PYTHIA and EPOS simulations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The b = -0.5 statistical baseline is not valid for fixed pT windows because the selected-particle fraction and pT variance change with Nch; the claimed dynamical signal may be a phase-space artifact.","rationale":"The reader's weakest assumption focused on whether pT cleanly separates soft and hard sources, and on the validity of the power-law form. The load-bearing issue I identify is more specific and precedes those: the paper's statistical baseline b = −0.5 is not applicable to fixed pT windows unless the fraction of particles in the window and their pT variance are independent of Nch. This is a quantitative, testable concern about the interpretation of the fitted index, not just about the physical meaning of pT regions. If the mixed-event control confirms a multiplicity-dependent acceptance effect, the paper's central claim—that deviations from −0.5 indicate dynamical correlations and a soft-hard crossover—does not follow from the shown data. The proceedings format is preliminary, and the ALICE collaboration typically performs such controls; therefore conditional acceptance is appropriate, with the mixed-event baseline as a required check before the interpretation is adopted.","tokens_in":5813,"tokens_out":4848,"duration_ms":45057,"concrete_test":"Construct a mixed-event or track-shuffling control: within each Nch class, randomly permute the pT values of tracks across events, preserving the per-event multiplicity and the class-averaged pT spectrum, then compute sqrt(Cm)/M(pT)m for the same pT windows and fit a*Nch^b over 10 < Nch < 70. If the mixed-event b values differ from −0.5 in the same direction and magnitude as the data (e.g., b ≈ −0.41 for 0.15–1.0 GeV/c), the claimed dynamical signal and soft-hard crossover are artifacts of the multiplicity-dependent pT acceptance. If the mixed-event b values are consistent with −0.5, the concern is resolved.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central conclusion rests on the statement in Section 2: 'In the absence of dynamical correlations, the correlator is expected to follow statistical scaling, yielding a power index of b = −0.5. Any deviation from this value would suggest the presence of dynamical correlations.' This expectation is only correct if the number of particles entering the correlator is proportional to Nch with a constant proportionality factor, and if the single-particle pT variance inside the chosen pT window is independent of Nch. Neither condition is established for the fixed pT windows used here. In pp collisions at 13 TeV the pT spectrum hardens with multiplicity, so for a window such as 0.15–1.0 GeV/c the fraction of accepted particles, and the variance of pT within that window, can change substantially with Nch. A pure superposition (no dynamical correlations) would then yield a power index b ≠ −0.5, with the deviation depending on the pT window. The observed b ≈ −0.41 for the lowest window and the non-monotonic behavior across windows can therefore be reproduced without any event-by-event dynamical correlation. The paper's interpretation of b deviations as evidence of 'dynamical correlations' and a 'crossover between soft and hard contributions' is thus not secure unless this trivial acceptance-induced scaling is controlled. The PYTHIA8 and EPOS LHC comparisons do not resolve the issue because both models also contain multiplicity-dependent pT spectra and are not a superposition baseline.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings-style paper reports an ALICE measurement of the event-by-event mean transverse momentum correlator sqrt(C_m)/M(pT)_m in pp collisions at sqrt(s)=13 TeV, studied as a function of charged-particle multiplicity and of the position/width of the pT selection window. The authors fit the multiplicity dependence with a power law a*Nch^b, quote values of b for several pT windows, interpret deviations from b=-0.5 as evidence for dynamical correlations, and compare the data with PYTHIA8 (Monash) and EPOS LHC. They conclude that the correlator is driven mainly by semi-hard and hard QCD processes, especially at high pT.","tokens_in":6030,"tokens_out":5754,"duration_ms":54842,"significance":"If the quantitative claims survive scrutiny, the paper would provide a useful differential extension of the earlier ALICE pp measurement, with new information on how the pT-correlation strength depends on the pT window. The comparison to PYTHIA8 and EPOS LHC is a sensible first step, and the power-law parametrization is a pragmatic way to summarize the trends. The paper is, however, built on a statistical baseline statement that is not derived and is, as written, not valid for fixed pT windows; this directly affects the central interpretation of the fitted index b. The absence of systematic uncertainties, fit-quality metrics, and experimental-selection details further prevents the reader from assessing the claimed deviations. The qualitative trends and the qualitative model comparison are valuable, but the quantitative conclusions are presently not supported.","major_comments":[{"comment":"The statement 'In the absence of dynamical correlations, the correlator is expected to follow statistical scaling, yielding a power index of b=-0.5' is not generally valid for the fixed pT windows used here. For independent particles, sqrt(C_m)/M(pT)_m scales as sqrt(sigma^2(Nch)/N_acc(Nch))/mu(Nch), where sigma and mu are the single-particle pT variance and mean inside the acceptance and N_acc is the number of accepted particles. If the pT spectrum hardens with multiplicity, the accepted fraction N_acc/N_ch and the within-window ratio sigma/mu can both depend on Nch; a pure superposition can then produce a power index different from -0.5, with a window-dependent deviation. The observed b≈-0.41 in the 0.15-1.0 GeV/c window and the non-monotonic dependence of b on window width may therefore be, at least in part, acceptance and spectral-shape effects rather than genuine event-by-event dynamical correlations. Please provide a derivation of the statistical baseline including these effects, or demonstrate quantitatively that they are negligible.","section":"Section 2 (after Fig. 2)"},{"comment":"The power-law fits are performed in the range 10<Nch<70 with the functional form a*Nch^b, but the paper gives no justification for this range or form and reports no goodness-of-fit (chi^2/ndf), number of fitted points, or systematic uncertainty on a and b. Since the central quantitative claim is the deviation of b from -0.5 and its variation with pT window, the manuscript should report the fit quality, stability with respect to the fit range, and the full uncertainty budget for b.","section":"Section 2 (power-law fits)"},{"comment":"The paper does not state the event-selection or track-selection criteria, trigger definitions, efficiency corrections, or the number of events used; the data points in Figures 1-4 are shown without systematic uncertainties. Without this information, the deviations of b and the quantitative agreement with PYTHIA and EPOS cannot be evaluated. At minimum, the authors should give the relevant ALICE analysis details and quote systematic uncertainties on the correlator values and on the fitted indices.","section":"Section 2 (experimental details)"},{"comment":"The comparison with PYTHIA8 and EPOS LHC is made only by visual inspection; the text states that agreement 'may be noticed to be rather better' at higher pT without any quantitative metric. Because both models themselves contain multiplicity-dependent pT spectra and are not a superposition baseline, they do not by themselves resolve the concern raised above about the b=-0.5 expectation. Please add a quantitative comparison (for example, ratio plots or chi^2 per point) and clarify what the model comparison demonstrates beyond the corrected statistical baseline.","section":"Section 2 (Figs. 3 and 4)"}],"minor_comments":[{"comment":"The phrase 'minimum bias and and high-multiplicity' contains a duplicated 'and'; please correct.","section":"Abstract"},{"comment":"The sentence 'In the case of widening pT windows, the strength of the correlator increases with multiplicity' appears to be a typo: the increase is with pT-window width, not with multiplicity, and the figure shows the correlator decreasing with Nch for each window.","section":"Section 2 (paragraph after Fig. 1)"},{"comment":"The equation is not typeset cleanly; the denominator 'Sigma_{nev,m} k N_pairs^k' is garbled and should be rewritten so that the pair-count normalization is unambiguous.","section":"Equation (1)"},{"comment":"The earlier ALICE points at sqrt(s)=7 TeV and Pb-Pb at 2.76 TeV are shown without uncertainties, and the quoted '~11% increase' from 7 to 13 TeV is not accompanied by a significance statement.","section":"Figure 2 right panel"},{"comment":"Reference [8] is a proceedings contribution by the author; the claim that a 'dedicated analysis' of high-multiplicity pp events revealed the decreasing trend should also cite the published ALICE paper [1].","section":"Section 1"},{"comment":"The sentence about 'non-monotonic variations in pT correlations with changing energy could serve as a signature of QGP formation' is not connected to any result in the paper; please either elaborate or remove it.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The paper is a proceedings-style contribution with preliminary ALICE figures. The main technical risk is the unvalidated b=-0.5 baseline: if the acceptance and spectral-shape effects are not controlled, the central interpretation collapses. I would encourage the editor to require the authors to either derive the corrected statistical baseline or soften the claims substantially. The self-citation [8] is unusual but not, by itself, a reason for concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a conference proceedings that extends the ALICE two-particle pT correlator measurement to 13 TeV pp and to wider and shifted pT windows (up to 6 GeV/c, and fixed-width windows from 1–2 through 4–5 GeV/c). The data are new and the figures are clear. The comparison with PYTHIA8 and EPOS LHC is useful and honestly presented. For that, credit is due. It is also plainly labeled ALICE Preliminary, which is the right framing for this format.\n\nThe soft spot is load-bearing. The paper claims that in the absence of dynamical correlations the correlator should scale as Nch^{-0.5}, and that fitted indices deviating from -0.5 are therefore evidence of dynamical correlations and a soft-to-hard crossover. That baseline is only valid if the number of accepted particles in the pT window is proportional to Nch with a constant fraction, and if the single-particle pT distribution inside the window does not change shape with multiplicity. Neither is established here. In pp at 13 TeV the spectrum hardens with multiplicity, so for windows like 0.15–1.0 GeV/c the selected fraction and the pT variance inside the window will vary with Nch. A trivial superposition would then produce an apparent power index different from -0.5, and the observed b ≈ -0.41 for the lowest window and the non-monotonic pattern across windows could be a phase-space artifact rather than a dynamical signal. The PYTHIA and EPOS comparisons do not rescue this, because both models also contain multiplicity-dependent pT spectra; they are not a superposition baseline. This needs to be controlled before the crossover interpretation can stand.\n\nOther issues are more minor. There are no systematic uncertainties, no data tables, no event-selection or track-efficiency details, and no fit-quality information – acceptable for a proceedings but it makes the numbers hard to evaluate. The abstract's phrase about \"signature of QGP formation\" overreaches for a single-energy pp measurement. Also, the text says the comparison to ref [1] was previously reported for pp at 13 TeV, but ref [1] is the 7 TeV measurement; the later sentence comparing 7 to 13 TeV contradicts the earlier mislabeling.\n\nWho is this for? Heavy-ion and QCD phenomenologists who want early look at ALICE's extended pT-fluctuation data. They will get some useful qualitative pictures, not a final result. A serious referee could help by forcing the acceptance-correction analysis and full systematics. For a conference proceedings, I would accept it with caveats; for a journal, major revision is needed. My recommendation: send it to peer review, but only if the referee insists on addressing the baseline issue.","headline":"New ALICE preliminary data on pT fluctuations at 13 TeV, but the claimed dynamical signal rests on an unjustified -0.5 baseline that likely breaks down for fixed pT windows.","tokens_in":6656,"tokens_out":2215,"would_cite":false,"duration_ms":21128,"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":"Event-by-event mean-transverse-momentum correlations in 13 TeV pp collisions decrease as a power of charged multiplicity, with an exponent that deviates from statistical independence and varies with the transverse-momentum window…","keywords":["event-by-event fluctuations","mean transverse momentum","two-particle pT correlator","multiplicity dependence","power-law scaling","soft-hard QCD crossover","proton-proton collisions","LHC"],"falsifier":"Recompute $\\sqrt{C_m}/M(p_{\\rm T})_m$ from mixed events—formed by drawing particles randomly from different events in the same multiplicity class so that single-particle spectra are preserved but inter-particle correlations are destroyed—and fit the same power law; if the measured index $b$ equals the mixed-event index (expected $-0.5$), then the dynamical deviations claimed in the paper are absent.","tokens_in":5538,"feed_emoji":"⚛️","tokens_out":10898,"duration_ms":80804,"temperature":0.7,"pith_summary":"This paper reports that event-by-event fluctuations of the mean transverse momentum in proton-proton collisions at 13 TeV shrink as the charged-particle multiplicity grows, and that the shrinkage follows a power law whose exponent differs from the pure-statistics value of -0.5. By moving the transverse-momentum acceptance upward, the correlation strength drops sharply and approaches zero, which the authors read as a crossover from soft processes (collective flow, resonance decays) to uncorrelated hard scatterings (jets, parton fragmentation). The fitted power-law index also depends on the width of the $p_{\\rm T}$ window in a non-monotonic way. The paper's conclusion is that mean-$p_{\\rm T}$ correlations in pp collisions are mostly driven by semi-hard and hard QCD processes, especially at high $p_{\\rm T}$. This matters because pp collisions are the baseline against which heavy-ion data are judged, and non-monotonic $p_{\\rm T}$ correlations across energies could be a signature of quark-gluon plasma formation.","feed_headline":"Power-law pT correlations expose soft-to-hard crossover in pp","feed_subtitle":"Event-by-event mean-pT correlations drop with multiplicity, signaling QCD correlations beyond statistical independence.","key_machinery":"The central object is the two-particle $p_{\\rm T}$ correlator $C_m$, defined for a multiplicity class $m$ as the average over events and over particle pairs of the product of each particle's deviation from the class-average mean $p_{\\rm T}$, normalized as $\\sqrt{C_m}/M(p_{\\rm T})_m$. The argument is carried by fitting this normalized correlator to $a\\cdot N_{\\rm ch}^{b}$ and comparing the exponent $b$ to the statistical superposition value $-0.5$; the systematic change of $b$ and of the correlation magnitude with the width and position of the $p_{\\rm T}$ window is what maps the soft-to-hard transition.","core_discovery":"The paper's central claim is that the normalized two-particle correlator $\\sqrt{C_m}/M(p_{\\rm T})_m$—the event-by-event fluctuation of the mean transverse momentum expressed in units of the average $p_{\\rm T}$—decreases with charged-particle multiplicity $N_{\\rm ch}$ in 13 TeV pp collisions, and that this decrease is described by a power law $a\\cdot N_{\\rm ch}^{b}$ over the range $10<N_{\\rm ch}<70$. The fitted index $b$ deviates from the expectation of $-0.5$ that would hold if particles were statistically independent, and the deviation depends on the position and width of the $p_{\\rm T}$ acceptance window: widening the window from 0.15–1.0 up to 0.15–6.0 GeV/$c$ raises the correlation strength, while shifting a fixed-width window from soft to hard $p_{\\rm T}$ suppresses it toward zero. The paper interprets this pattern as a crossover between soft and hard sources, and notes that a pQCD-based Monte Carlo model captures the high-$p_{\\rm T}$ behavior better than a core-corona model. The conclusion asserts that mean-$p_{\\rm T}$ correlations in pp collisions are mainly driven by semi-hard and hard QCD processes, particularly at high $p_{\\rm T}$.","pith_inferences":["A sharper test of the claimed crossover would be to map $b$ as a function of the upper edge of the $p_{\\rm T}$ window in fine steps; a sharp transition at some characteristic $p_{\\rm T}$ would locate the momentum scale where jet contributions overcome soft ones.","The same analysis could be run with mixed events that preserve single-particle spectra but destroy inter-particle correlations; any measured deviation of $b$ from the mixed-event value would then quantify the dynamical-correlation fraction directly, which the paper does not subtract.","The model comparison would be more decisive if the power-law index $b$ were extracted from the Monte Carlo predictions in the same $10<N_{\\rm ch}<70$ fit range and compared numerically, rather than only by eye.","If the soft-to-hard crossover is real, the normalized correlator for identified particles of different masses might separate decay-driven correlations from genuine flow-like correlations in the low-$p_{\\rm T}$ window."],"forward_implications":["If the conclusion is right, the low-$p_{\\rm T}$ window (0.15–1.0 GeV/$c$) carries most of the dynamical correlation, so future measurements comparing pp with heavy-ion collisions must match the acceptance window before interpreting differences as collective effects.","The deviation of $b$ from $-0.5$ gives a quantitative, multiplicity-dependent observable that any Monte Carlo generator must reproduce; the better agreement of the pQCD model at wide $p_{\\rm T}$ windows makes hard QCD the likely driver in that regime.","High-multiplicity pp events cannot serve as a purely soft baseline; their mean-$p_{\\rm T}$ correlations already contain a jet-induced component that grows with the $p_{\\rm T}$ acceptance.","If $b$ indeed approaches the statistical value at very high $p_{\\rm T}$, then the high-$p_{\\rm T}$ window isolates approximately independent hard scatterings, providing a practical way to measure the hard component separately."],"supporting_citations":[{"why":"Defines the two-particle $p_{\\rm T}$ correlator and supplies the earlier 7 TeV pp and 2.76 TeV Pb-Pb measurements whose power-law exponents and superposition expectation are the paper's baseline.","marker":"[1]"},{"why":"Establishes that hydrodynamical collectivity dominates the low-$p_{\\rm T}$ region, underpinning the soft-versus-hard interpretation of the $p_{\\rm T}$ windows.","marker":"[5]"},{"why":"Supplies the picture that low-$p_{\\rm T}$ correlations reflect collective flow and resonance decays, used to explain the drop in correlation strength when the window is shifted upward.","marker":"[16]"},{"why":"Provides the PYTHIA8 generator whose pQCD-based description of hard scattering gives the best agreement with data at wide and high-$p_{\\rm T}$ windows.","marker":"[20]"},{"why":"Provides the EPOS LHC core-corona model that underestimates the correlation strength, serving as the contrasting prediction.","marker":"[21]"}],"fun_headline_variants":["pT fluctuation power law signals soft-hard crossover in pp","Power-law drop in pT fluctuations maps pp soft-to-hard transition","Multiplicity-driven power law in pT correlations reveals soft-hard mix","Power-law pT fluctuations across multiplicity hint at QCD sources","Soft-hard interplay in pp seen in power-law pT fluctuation drop"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumption that the chosen power-law form $aN_{\\rm ch}^b$ and the fit range $10<N_{\\rm ch}<70$ describe the data, and that $p_{\\rm T}$ cleanly separates soft from hard particle production; if either fails, the reported exponent and its physical interpretation do not follow.","fun_headline_variants_meta":{"raw":{"variants":["pT fluctuation power law signals soft-hard crossover in pp","Power-law drop in pT fluctuations maps pp soft-to-hard transition","Multiplicity-driven power law in pT correlations reveals soft-hard mix","Power-law pT fluctuations across multiplicity hint at QCD sources","Soft-hard interplay in pp seen in power-law pT fluctuation drop"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000694,"raw_usage":{"total_tokens":3220,"prompt_tokens":1109,"completion_tokens":2111,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":725,"completion_tokens_details":{"reasoning_tokens":2024}},"tokens_in":725,"tokens_out":2111,"duration_ms":15011,"temperature":1.0,"reasoning_tokens":2024,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T15:22:24.247931+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute $\\sqrt{C_m}/M(p_{\\rm T})_m$ from mixed events—formed by drawing particles randomly from different events in the same multiplicity class so that single-particle spectra are preserved but inter-particle correlations are destroyed—and fit the same power law; if the measured index $b$ equals the mixed-event index (expected $-0.5$), then the dynamical deviations claimed in the paper are absent.","supporting_citations":[{"cited_title":"Abelev et al, (ALICE Collaboration) Euro","cited_arxiv_id":null,"evidence_quote":"Defines the two-particle $p_{\\rm T}$ correlator and supplies the earlier 7 TeV pp and 2.76 TeV Pb-Pb measurements whose power-law exponents and superposition expectation are the paper's baseline."},{"cited_title":"Adler et al, (STAR Collaboration) Phys","cited_arxiv_id":null,"evidence_quote":"Establishes that hydrodynamical collectivity dominates the low-$p_{\\rm T}$ region, underpinning the soft-versus-hard interpretation of the $p_{\\rm T}$ windows."},{"cited_title":"Acharya et al, (ALICE Collaboration) Phys","cited_arxiv_id":null,"evidence_quote":"Supplies the picture that low-$p_{\\rm T}$ correlations reflect collective flow and resonance decays, used to explain the drop in correlation strength when the window is shifted upward."},{"cited_title":"Sj¨ ostrand et al., Comput","cited_arxiv_id":null,"evidence_quote":"Provides the PYTHIA8 generator whose pQCD-based description of hard scattering gives the best agreement with data at wide and high-$p_{\\rm T}$ windows."},{"cited_title":"Pierog et al., Phys","cited_arxiv_id":null,"evidence_quote":"Provides the EPOS LHC core-corona model that underestimates the correlation strength, serving as the contrasting prediction."}],"review_version":1}