{"id":"dbe34a0c-08a9-4727-be7a-bebe7a04dd06","arxiv_id":"2412.21116","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The parton branching TMD framework is upgraded from NLL to NNLL accuracy using the soft-gluon physical coupling, with the Collins-Soper kernel evaluated at NNLL.","lead":"This paper extends the parton branching method for transverse momentum dependent (TMD) quark and gluon distributions to next-to-next-to-leading-logarithm (NNLL) accuracy by adopting the soft-gluon physical coupling. The result matters because NNLL is the first order sensitive to the collinear anomaly and could improve precision predictions for LHC observables such as Z boson spectra.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The NNLL claim rests on an unproven, imported relation: Eq. (13) is asserted without a derivation from Eq. (2), and Eq. (16) is explicitly delegated to the authors' unpublished manuscript [65], so the central claim is not independently verifiable from the text.","rationale":"The Reader's conditional verdict already identifies the two critical weaknesses: the soft-coupling replacement is imported from Refs. [47,48] rather than re-derived, and Eq. (16) depends on the unpublished Ref. [65]. I agree with both points and with CONDITIONAL as the appropriate verdict. My stress-test sharpens the first point by noting a specific potential omission that the Reader did not name: when α_s → α_s^phys is inserted into Eq. (2), the d_a coefficient is also transformed, and at O(α_s^2) this can produce single-logarithmic terms alongside the quoted d_a^(1). The paper subtracts K^(1) to avoid double counting the cusp part, but does not show how the accompanying K^(1)d^(0) term is handled. If that term is in fact absent or absorbed into the chosen MS resummation scheme, the claim survives; if it is not, the advertised NNLL B coefficient would be incomplete. This is a concrete, checkable algebraic question. The second concern, Eq. (16), is a verifiability issue more than a proven error; it can be settled by an independent comparison with published NNLL results. Neither concern is strong enough to reject the claim on the evidence at hand, since the ingredients are standard and the numerical outputs are plausible. The conditionality of the Reader's verdict is therefore exactly right: the paper's central claim should be accepted only once the derivation and Eq. (16) are independently available and checked. I set verdict_should_be to UNCHANGED to indicate that no change to the Reader's CONDITIONAL verdict is needed.","tokens_in":11430,"tokens_out":17780,"duration_ms":182791,"concrete_test":"Independently derive Eq. (13) from Eq. (2) by inserting Eq. (9) and the two-loop k_a,d_a of Eqs. (6)–(8), and by performing the z and μ' integrals through O(α_s^3), including all terms generated from the d_a(α_s^phys) contribution. Then compare the resulting exponent term-by-term with Eq. (13). A second check: recompute Eq. (16) using published O(α_s^3) cusp-anomalous-dimension and Collins-Soper kernel coefficients; if the numerical value of A_a^(3) − k_a^(2) does not match, the asserted link between the PB NNLL Sudakov form factor and the CS kernel is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central assertion—that NNLL accuracy is reached by substituting the soft-gluon physical coupling into the PB Sudakov form factor—depends on two load-bearing steps that the manuscript does not establish. First, Eq. (13) is introduced as the result of 'performing the longitudinal momentum integration,' but the derivation is not shown. In particular, the effect of the replacement α_s → α_s^phys on the d_a(α_s) term is not displayed: expanding Eq. (9) to O(α_s^2) generates additional single-logarithmic contributions (for example a K^(1)d^(0) term if the same physical coupling multiplies d_a), and the text does not demonstrate that these are cancelled, absorbed, or genuinely absent from the quoted B_a^(2) = −2 d_a^(1). Second, Eq. (16), which is the only relation connecting the computed A_a^(3) − k_a^(2) to the Collins-Soper kernel, is attributed to the authors' own unpublished paper [65]; no independent derivation or cross-check is supplied. Since Eq. (13) and Eq. (16) are the two places where 'NNLL' is realized, both unverified links are load-bearing. The result may well be correct—the ingredients are documented and the numerics are illustrative—but as written the manuscript does not allow a reader to verify the advertised accuracy.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript claims that next-to-next-to-leading-logarithm (NNLL) accuracy can be achieved in the parton-branching (PB) TMD framework by replacing the strong coupling with the soft-gluon physical coupling. The central result is Eq. (13), which gives the perturbative PB Sudakov form factor with double-logarithmic coefficient A_a^(3)=K^(2)k_a^(0) and single-logarithmic coefficient B_a^(2)=-2d_a^(1), together with the associated identification of the Collins-Soper kernel through Eq. (16). The paper presents numerical illustrations for TMD distributions, the Z-boson transverse momentum spectrum, and the b-dependence of the Collins-Soper kernel, comparing with several extractions from the literature. The manuscript explicitly states that full details will be reported elsewhere [65], and Eq. (16) is attributed to the authors' own unpublished manuscript.","tokens_in":11774,"tokens_out":10203,"duration_ms":96405,"significance":"If the central derivation is correct, this would be the first NNLL computation with PB TMD techniques and would open the way to improved PB-based predictions for LHC observables. The paper usefully assembles known ingredients: the soft-gluon physical coupling, two-loop splitting functions, the dynamical resolution scale, and the CS-kernel extraction technique. The numerical comparisons in Figs. 1 and 2 are transparent and the relation to existing TMD extractions is clearly presented. However, the advertised NNLL result is not independently verifiable from the text: Eq. (13) is asserted without showing the integration steps, and Eq. (16) is delegated to an unpublished companion paper. The paper does not ship machine-checked proofs or reproducible code; the numerical results are illustrative, with scale-variation bands only.","major_comments":[{"comment":"Equation (13) is the central result, but the derivation from Eq. (2) is not shown. The statement 'By performing the longitudinal momentum integration' skips the expansion of the physical coupling in the Sudakov exponent. In particular, the manuscript does not demonstrate that substituting α_s^phys into the d_a(α_s) term produces no additional single-logarithmic contributions at the order of interest, nor does it show explicitly how the subtracted coupling α_s^subtr removes all K^(1)-induced terms beyond the double-counting of k_a^(1). Without these steps, Eq. (13) is an assertion rather than a derived result, and the identification B_a^(2)=-2d_a^(1) is not independently established. Please provide the full derivation or a pointer to a derivation in a published, accessible reference.","section":"§2, text preceding Eq. (13)"},{"comment":"Equation (16) is the only relation connecting the computed NNLL double-logarithmic coefficient A_a^(3) to the Collins-Soper kernel, and it is attributed to the authors' own unpublished manuscript [65]. The sentence 'Full details will be reported elsewhere [65]' in the introduction likewise leaves the advertised NNLL result unverifiable. Because Eq. (16) is load-bearing for the CS-kernel results in Fig. 2, the manuscript should either derive this relation explicitly or replace reference [65] with a published or publicly available derivation. As written, a reader cannot check the consistency of A_a^(3), k_a^(2), and the CS kernel.","section":"§4, Eq. (16)"},{"comment":"The circularity risk in the claim A_a^(3)=K^(2)k_a^(0) should be addressed explicitly. The coefficient K^(2) in Eq. (11) is defined in Refs. [47,48] precisely through the requirement that the physical coupling reproduces NNLL resummation in soft-gluon radiation. The manuscript should clarify what is newly derived in the PB framework versus what is imported from Refs. [47,48], and should specify why no additional PB-specific corrections enter the NNLL coefficient. A concrete check against a known NNLL result in a limiting case (for example, matching Eq. (13) to the standard resummed Sudakov exponent at O(α_s^3)) would considerably strengthen the claim that this is a genuine PB computation of NNLL accuracy.","section":"§2, Eqs. (9)–(13)"}],"minor_comments":[{"comment":"The typography of Eq. (13) makes the power counting ambiguous: in the third line, the factor α_s/(2π) multiplying d_a^(1) appears to sit at a different order in α_s than the preceding K^(2)k_a^(0) term. Please clarify the intended bracket structure and the overall α_s power of each term.","section":"Eq. (13)"},{"comment":"The phrase 'with the explicit expressions given in Eqs. (7),(8)' is confusing because Eqs. (7) and (8) give the d_a^(1) coefficients, while the scheme-difference relations in Eqs. (14) and (15) involve B_a^(2). Please state explicitly which scheme the PB d_a coefficients are defined in before comparing with the DY and Higgs schemes.","section":"§2, text after Eq. (15)"},{"comment":"The caption refers to 'the top five entries in the legend' but does not list the five scenarios. Please enumerate them (for example, in the caption or in a table) so that the curves can be identified without relying on the legend alone.","section":"Fig. 2 caption"},{"comment":"The text says 'The PB result for the rapidity-evolution kernel is presented next,' but no explicit formula for the kernel is given; only Eq. (16) and Fig. 2 appear. Please state the final expression for the CS kernel, including the non-perturbative contribution from region b), or indicate explicitly that this is deferred to Ref. [65].","section":"§4, sentence before Eq. (16)"},{"comment":"Reference [65] is listed as 'in preparation, (2024)'; if it remains unpublished, the manuscript should not rely on it for the central relation Eq. (16). Please update the citation or provide the derivation in the text.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's central claim is potentially important, but the current text defers the key derivation and a load-bearing relation to an unpublished companion paper [65]. For a journal publication, this is a significant verifiability problem. The authors should be asked to include the derivation of Eq. (13) and of Eq. (16) (or to cite a published source for the latter), and to discuss explicitly the sense in which A_a^(3)=K^(2)k_a^(0) is a PB prediction rather than a restatement of the definition of the physical coupling. I do not see a fatal internal inconsistency, so major revision seems appropriate rather than rejection, provided the missing material can be supplied."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a credible announcement of the first NNLL-accurate PB TMD Sudakov form factor, but as a standalone paper it does not show its own derivation. The central result, Eq. (13), appears after a one-line 'performing the longitudinal momentum integration,' and the relation to the Collins-Soper kernel, Eq. (16), is credited to the authors' own unpublished manuscript [65]. The paper itself says 'Full details will be reported elsewhere.' So the claim is plausible but not independently verifiable from the text.\n\nWhat is genuinely new: they upgrade the PB TMD framework from NLL to NNLL, and they identify that the NNLL double-log coefficient A^(3) comes from the soft-gluon physical coupling rather than from three-loop splitting functions, consistent with the collinear anomaly. The connection between the difference A^(3) - k^(2) and the CS kernel is a new structural observation in this scheme. The numerical illustrations (Fig. 1 for the DY spectrum, Fig. 2 for the CS kernel against literature extractions) are just illustrations, but they show the machinery works and the comparison to lattice and data extractions is a useful sanity check.\n\nThe soft spots, in proportion: the biggest is the deferred derivation. Eq. (13) and Eq. (16) are the two load-bearing places where 'NNLL' actually appears, and both are taken on faith. The stress-test note worries about an extra K^(1)d^(0) term when substituting the physical coupling into the d_a term; on reading, the paper says they use the subtracted coupling α_s^subtr = α_s^phys - K^(1)α_s^2/(2π), so that specific term is removed by construction. But the paper does not show the cancellation of other scheme-dependent artifacts, and the overall derivation is absent. The circularity concern about A^(3)=K^(2)k^(0) is real but mild: the coefficient is, by construction, the one encoded in the physical coupling; the informative part is that the PB scheme keeps it intact and connects it to the CS kernel. The numerics have only scale-variation bands, not a full uncertainty treatment, but they are explicitly labeled as illustrative.\n\nMy bottom line: this deserves peer review. It's a within-subfield advance with real consequences for LHC precision, and the authors are transparent about what is deferred. A referee should be able to check the consistency of Eq. (13) against the conventional resummation scheme in Refs. [76,77] and ask for the companion paper or the missing steps before endorsing the NNLL claim. I'd send it to review, with the clear expectation that the derivation be supplied or the companion released.\n\nFor your reading group: maybe. It's short and the plot comparing CS kernels is thought-provoking.","headline":"Plausible and important first NNLL step for PB TMD, but the derivation is deferred to an unpublished companion, so treat the claim as conditional.","tokens_in":12339,"tokens_out":9057,"would_cite":false,"duration_ms":75881,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["12.38.-t","12.38.Bx"],"model":"deepseek-v4-flash","headline":"By substituting the soft-gluon physical coupling into the parton-branching Sudakov form factor, the paper achieves NNLL accuracy for TMD evolution and evaluates the Collins-Soper kernel at that order.","keywords":["transverse momentum dependent distributions","parton branching","Sudakov form factor","soft-gluon physical coupling","NNLL resummation","Collins-Soper kernel","Drell-Yan production","QCD evolution"],"falsifier":"Compute the PB Sudakov form factor directly at three loops without the coupling substitution and check whether the order-$\\alpha_s^3$ double-logarithmic coefficient equals $K^{(2)} k_a^{(0)}$; if it does not, the soft-gluon coupling replacement is incomplete for this framework. A second decisive check is to extract the Collins-Soper kernel from PB TMD predictions at small transverse coordinate and compare with lattice determinations, since a mismatch would falsify Eq. (16).","tokens_in":11264,"feed_emoji":"⚛️","tokens_out":13962,"duration_ms":118307,"temperature":0.7,"pith_summary":"This paper aims to establish that transverse-momentum-dependent (TMD) parton distributions in the parton-branching (PB) approach can be evolved at next-to-next-to-leading-logarithmic (NNLL) accuracy, a step beyond the NLL accuracy previously reached in this framework. The central step is to replace the ordinary QCD strong coupling in the PB Sudakov form factor with the soft-gluon physical coupling, the higher-order extension of the classic one-loop soft-coupling result. This replacement produces the explicit NNLL coefficients, double-logarithmic $A^{(3)} = K^{(2)} k^{(0)}$ and single-logarithmic $B^{(2)} = -2 d^{(1)}$, in the perturbative Sudakov form factor. The paper also evaluates the Collins-Soper kernel at NNLL, including nonperturbative contributions, and gives illustrative Drell-Yan $Z$-boson transverse momentum predictions. If the claim holds, PB-based TMD evolution and parton-shower Monte Carlo predictions can be upgraded to NNLL accuracy for precision collider physics.","feed_headline":"Soft-gluon coupling lifts parton-branching TMDs to NNLL accuracy","feed_subtitle":"New NNLL Sudakov coefficients sharpen Drell-Yan predictions and the rapidity-evolution kernel at LHC precision","key_machinery":"The central object is the soft-gluon physical coupling, $\\alpha_s^{\\rm phys} = \\alpha_s\\bigl(1 + \\sum_{n\\ge 1} K^{(n)} (\\alpha_s/2\\pi)^n\\bigr)$, with $K^{(1)}$ the classic one-loop coefficient and $K^{(2)}$ taken from higher-order soft-gluon resummation. Inserting this coupling in its subtracted form, $\\alpha_s^{\\rm subtr} = \\alpha_s^{\\rm phys} - K^{(1)}\\alpha_s^2/(2\\pi)$, into the PB Sudakov form factor, together with the angular-ordering relation $q_\\perp = (1-z)\\mu'$ that splits the branching integral into perturbative and nonperturbative regions, yields the NNLL coefficients of Eq. (13). The same setup supplies the identity connecting the NNLL double-log coefficient to the Collins-Soper kernel, Eq. (16), which the paper uses to compute the kernel at NNLL.","core_discovery":"The paper's claim, stated on its own terms, is that the PB Sudakov form factor evaluated with two-loop splitting functions and the subtracted soft-gluon physical coupling is NNLL accurate. The resulting form factor, Eq. (13), has an $O(\\alpha_s^3)$ double-logarithmic coefficient $K^{(2)} k_a^{(0)} \\equiv A_a^{(3)}$ supplied by the soft-gluon coupling, and an NNLL single-logarithmic coefficient $-2 d_a^{(1)} \\equiv B_a^{(2)}$ supplied by the two-loop splitting functions. The paper further observes that at NNLL the double-log coefficient is no longer proportional to the cusp anomalous dimension because of the collinear anomaly, and it uses Eq. (16) to relate the difference $A_a^{(3)} - k_a^{(2)}$ to the perturbative Collins-Soper kernel. It then computes the Collins-Soper kernel at NNLL, adding a nonperturbative contribution from the large-distance region, and compares with data-driven and lattice extractions. The authors describe this as the first NNLL computation performed with PB TMD techniques.","pith_inferences":["Beyond the paper: the recipe of replacing the coupling by its physical soft-gluon form may transfer to other parton-shower evolutions, suggesting that NNLL accuracy can be gained without rebuilding the splitting functions from scratch.","Beyond the paper: the flattening of the Collins-Soper kernel at large $b$ seen for the dynamical resolution scale, if confirmed by data, would support saturation-like behavior of the kind preferred by recent Drell-Yan fits.","Beyond the paper: a dedicated fit of the NNLL PB TMD predictions to precise $Z$ $p_T$ data could pin down the nonperturbative resolution scale $q_0$ and discriminate between $\\alpha_s(q_\\perp)$ and $\\alpha_s(\\mu)$ coupling prescriptions.","Beyond the paper: since NNLL is the first order at which the collinear anomaly enters, the low-$p_T$ region of the $Z$ spectrum is a direct place to test whether the PB framework really captures the physics at this order."],"forward_implications":["PB TMD distributions can be evolved at NNLL accuracy in the Sudakov region, directly improving predictions for low-$p_T$ Drell-Yan observables at the LHC.","The Collins-Soper kernel becomes computable in the PB framework at NNLL, with nonperturbative large-$b$ contributions, allowing direct comparison with data-driven and lattice results.","The NNLL Sudakov form factor can be combined with existing NLO matching and multi-jet merging machinery to produce more accurate Monte Carlo predictions.","At NNLL the double-logarithmic Sudakov coefficient explicitly separates from the cusp anomalous dimension through the collinear anomaly, and the PB framework makes that separation concrete through the Collins-Soper kernel.","Because the soft-gluon coupling is known beyond $O(\\alpha_s^3)$, the same construction can in principle be extended systematically to higher logarithmic orders."],"supporting_citations":[{"why":"It defines the PB TMD evolution framework and supplies the $k_a$, $d_a$ splitting-function coefficients used in the Sudakov form factor.","marker":"[33, 34]"},{"why":"It supplies the second-order soft-gluon coupling coefficient $K^{(2)}$ that generates the NNLL double-logarithmic term.","marker":"[47, 48]"},{"why":"It gives the first-order soft-gluon coupling and the angular-ordering treatment of soft radiation that underlies the branching kinematics.","marker":"[49]"},{"why":"It identifies the collinear anomaly that makes the NNLL double-log coefficient differ from the cusp anomalous dimension.","marker":"[50]"},{"why":"It derives the angular-ordering relation $q_\\perp=(1-z)\\mu'$ and the split of the Sudakov region into perturbative and nonperturbative parts.","marker":"[70]"},{"why":"It provides the ratio technique used to extract the Collins-Soper kernel from Drell-Yan transverse momentum distributions.","marker":"[53]"},{"why":"It is the companion paper from which Eq. (16), relating $A^{(3)}-k^{(2)}$ to the Collins-Soper kernel derivative, is taken.","marker":"[65]"}],"fun_headline_variants":["Soft-gluon coupling yields NNLL parton-branching TMDs","First NNLL parton-branching TMD via soft-gluon coupling","Soft-gluon coupling delivers NNLL Sudakov for TMDs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands on the assumption that substituting the physical soft-gluon coupling for the ordinary strong coupling inside the PB Sudakov form factor captures the complete NNLL double-logarithmic correction with no further parton-branching-specific terms, and on a companion paper (currently unpublished) for the identity linking that coefficient to the Collins-Soper kernel.","fun_headline_variants_meta":{"raw":{"variants":["Soft-gluon coupling yields NNLL parton-branching TMDs","First NNLL parton-branching TMD via soft-gluon coupling","Soft-gluon coupling delivers NNLL Sudakov for TMDs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000539,"raw_usage":{"total_tokens":2553,"prompt_tokens":877,"completion_tokens":1676,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":493,"completion_tokens_details":{"reasoning_tokens":1610}},"tokens_in":493,"tokens_out":1676,"duration_ms":10337,"temperature":1.0,"reasoning_tokens":1610,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:01:34.963049+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the PB Sudakov form factor directly at three loops without the coupling substitution and check whether the order-$\\alpha_s^3$ double-logarithmic coefficient equals $K^{(2)} k_a^{(0)}$; if it does not, the soft-gluon coupling replacement is incomplete for this framework. A second decisive check is to extract the Collins-Soper kernel from PB TMD predictions at small transverse coordinate and compare with lattice determinations, since a mismatch would falsify Eq. (16).","supporting_citations":[{"cited_title":"Extending parton branching TMDs to small $x$","cited_arxiv_id":"1908.01621","evidence_quote":"It gives the first-order soft-gluon coupling and the angular-ordering treatment of soft radiation that underlies the branching kinematics."},{"cited_title":"Soft-gluon effective coupling: perturbative results and the large-nF limit to all orders","cited_arxiv_id":"2309.11584","evidence_quote":"It identifies the collinear anomaly that makes the NNLL double-log coefficient differ from the cusp anomalous dimension."}],"review_version":1}