{"id":"beb4084b-fad1-41ab-917b-3a8feb850b05","arxiv_id":"1908.03831","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors compute NNLO quark transverse momentum dependent PDFs and fragmentation functions with the exponential regulator, and extract the two-loop quark jet function needed for N3LL energy-energy correlator resummation.","lead":"This paper computes a key set of quantum chromodynamics ingredients that describe how quarks with sideways momentum are distributed inside hadrons and inside jets. It also provides the last missing two-loop piece for a very precise prediction of the energy-energy correlator in the back-to-back limit.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The zero-bin subtraction in Eq. (2.27) is load-bearing: the TMDFF discrepancy with Ref. [40] sits exactly at the identification of the zero-bin with the TMD soft function, and the claimed EEC constant cJ2 inherits this assumption.","rationale":"The reader identified Eq. (2.27) as a weak assumption, and I agree it is the most load-bearing point. The paper is careful, with strong consistency checks, and the TMDPDF agreement with Refs. [38-40] independently validates the exponential regulator in the collinear sector. However, the TMDFF discrepancy is the one unresolved spot, and the paper's own resolution of it—the jet-function check—does not fully isolate the disputed constant. The EEC jet function check is a moment of the TMDFF sum, so it validates the combination C_qq + C_gq + C_barqq + 2(Nf-1) C_q'q, not each individually. The N=4 check only probes the maximal transcendental part; cJ2 in Eq. (4.16) contains subleading transcendental pieces like -178/3 CA CF zeta2 that are not checked. The sum rule checks validate the delta(1-z) coefficient of the physical cross section but again only in combination with hard and soft functions. Thus the concern is not that the result is wrong, but that a specific scheme-consistency assumption is both load-bearing and under-verified at exactly the order where the discrepancy appears. The concrete test I propose would settle it; until then CONDITIONAL is the right verdict, matching the reader. I do not think the concern rises to REJECT because there is no internal inconsistency offered, and because the authors' checks are substantial independent evidence.","tokens_in":31757,"tokens_out":2069,"duration_ms":21632,"concrete_test":"Recompute the q-to-q TMDFF zero-bin contribution directly in the exponential scheme using the operator definition Eq. (2.26) with the zero-bin limit taken both before and after the tau-to-0 limit, and compare against S_qbarq of Eq. (2.24) at O(epsilon^2). Specifically, evaluate the two-loop soft region of the TMDFF collinear matrix element with the exponential regulator and confirm it equals the TMD soft function to all orders in epsilon. If it differs, insert the corrected zero-bin into Eq. (2.29) and recompute cJ2 in Eq. (4.16); a change in the CACF pi^4 coefficient would resolve the Ref. [40] discrepancy.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim—that Eq. (4.16) is the last missing two-loop ingredient for N3LL EEC resummation—is downstream of the TMDFF matching coefficients, which rest on the assertion in Eq. (2.27) that the zero-bin subtraction equals the TMD soft function S0b = S_qbarq. This identification is stated rather than derived. If the zero-bin soft region receives a different τ-regularized contribution in the collinear TMDFF sector than in the soft-function definition (e.g., a different subleading-in-τ region or an overlap missed because the regulator multiplies the total k0), then the O(epsilon^2) bare results and the constant cJ2 shift. The paper itself reports a small discrepancy with Ref. [40] in the q-to-q channel, specifically a CACF pi^4 delta(1-z) term that it attributes to the universal TMD soft function; this is precisely the type of term whose coefficient is controlled by the zero-bin subtraction. Since the Ref. [40] comparison uses a different (delta) regulator, the disagreement signals that the exponential regulator's zero-bin identification may not be scheme-independent. The checks listed (RGE structure, N=4 leading-transcendental check, energy/momentum sum rules) all constrain the sum over channels or the leading transcendental part, but the disputed constant term is a color-mixed, subleading-transcendental quantity that those checks do not uniquely pin down.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper revisits the NNLO calculation of quark transverse momentum dependent parton distribution functions (TMDPDFs) and fragmentation functions (TMDFFs) using the exponential regulator for rapidity divergences. The authors show that the regulator can be applied at the operator level, simplifies the phase-space integrals, and enables the use of IBP and differential-equation techniques. They present the scale-independent NNLO matching coefficients, report agreement with existing TMDPDF results, and find a discrepancy with Ref. [40] in the q-to-q TMDFF channel. As a by-product, they extract the two-loop quark jet function for the energy-energy correlator (EEC) in the back-to-back limit, including the new constant cJ2, and claim this is the last missing ingredient for N3LL resummation.","tokens_in":32023,"tokens_out":7013,"duration_ms":83552,"significance":"If the results are correct, the paper provides a valuable technical step: it demonstrates that the exponential regulator works for collinear TMD sectors, supplies O(epsilon^2) bare results needed for future N3LO calculations, and delivers a previously missing two-loop ingredient for the EEC. The calculation uses modern and reliable tools (IBP reduction, canonical differential equations, HPLs), and the paper includes several internal consistency checks: reproduction of RGE and rapidity-evolution structures, agreement with prior TMDPDF calculations, a leading-transcendentality check against N=4 SYM, and a momentum-conservation sum rule. These strengths are substantial. However, the checks do not fully pin down the disputed constant term, and one load-bearing assumption is stated rather than derived.","major_comments":[{"comment":"The identification of the zero-bin subtraction with the full TMD soft function S_qbarq is stated without derivation. This equality is load-bearing because the disputed q-to-q TMDFF term C_A C_F pi^4 delta(1-z) in Sec. 4.1 is attributed precisely to the universal soft function. The exponential regulator is implemented differently in Eqs. (2.24)-(2.26): a coordinate shift for the collinear matrix elements versus a Wilson-loop definition for the soft function, and the zero-bin is the soft limit of the collinear graph rather than an independently defined object. Please provide a derivation or an explicit diagram-by-diagram check that the soft limit of each collinear cut diagram equals the corresponding soft-function contribution with the same tau prescription, including subleading-in-tau regions. If this identification fails in any channel, the O(epsilon^2) bare results and the constant cJ2 in Eq. (4.16) are scheme-dependent.","section":"Sec. 2.2, Eq. (2.27)"},{"comment":"The statement that the two-loop plus-distribution terms D_n(z) are in full agreement with the 'analytical NLO calculation' in Ref. [82] is difficult to interpret, because Ref. [82] is a one-loop calculation and cannot determine O(alpha_s^2) D_n coefficients. Please clarify what comparison was actually performed, for example whether the two-loop leading singular terms were checked against an independent factorization-based calculation rather than against the NLO result. As written, this statement appears to claim a check that the cited calculation cannot provide.","section":"Sec. 4.2, Eq. (4.17)"},{"comment":"The claimed verification of the momentum-conservation sum rule does not test the two-loop endpoint contribution. The analytical NLO formula from Ref. [82] supplies only the one-loop cross section, while Eq. (4.17) contains the two-loop endpoint term dsigma^(2)/dz. A two-loop verification of the sum rule would require the full two-loop EEC away from the endpoint or an independent derivation of the endpoint integral. As presented, this check constrains only lower orders and leaves cJ2 in Eq. (4.16) and c_{z=1}^{(2)} in Eq. (4.19) without an independent numerical or analytical test beyond the partial N=4 leading-transcendentality check.","section":"Sec. 4.2, Eq. (4.20)"},{"comment":"The claim that the loop integral in the real-virtual contribution 'does not need to be regularized' is not demonstrated. The master integrals IRV_1 and IRV_2 in Eq. (3.15) contain factors [nbar dot l]^{-a4}, and the cancellation of rapidity singularities between the real and virtual contributions is a nontrivial consistency requirement when the exponential regulator acts only on the real-emission phase space. Please show explicitly, or cite a derivation, that the combined real-virtual plus double-real contributions are finite in the tau -> 0 limit before zero-bin subtraction.","section":"Sec. 3.2.1, Eqs. (3.15)"}],"minor_comments":[{"comment":"The quantity L_nu = ln(nu^2/mu^2) is used in Eq. (3.9) but is not defined until the appendix; please define it near Eq. (2.38) for readability.","section":"Sec. 3.1, Eq. (3.9)"},{"comment":"There is a typo in the sentence 'probability distribution of find a quark'; it should read 'of finding a quark'.","section":"Sec. 2.1"},{"comment":"The phase-space regulator in Eq. (2.22) uses exp(-b0 tau k0) with a real exponent, while the operator definitions in Eqs. (2.24)-(2.26) use a shift by -i b0 tau in spacetime arguments. Please clarify the relationship between these two representations and the Euclidean-signature convention for the transverse components.","section":"Sec. 2.2, Eqs. (2.22) and (2.24)"},{"comment":"The O(epsilon^2) bare results are said to be provided in an electronic file attached to the arXiv submission, but the file format and the precise definitions of the plotted or listed quantities are not described in the text. A short summary of the file contents would help readers use these results.","section":"Sec. 3.2.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically serious and the authors are well known in the field, but the missing derivation of the zero-bin identity and the questionable EEC checks are exactly the points on which the disputed TMDFF result and the new constant cJ2 rest. The authors should be asked to provide a detailed derivation or a dedicated numerical/analytic cross-check of the zero-bin subtraction, and to correct the claims about the checks in Sec. 4.2. These are fixable within the scope of the manuscript, so I do not recommend rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper is worth a serious referee, and I'd be inclined to accept it for review despite one unresolved discrepancy. The genuinely new pieces are the O(epsilon^2) bare results for the quark TMDPDF/TMDFF with the exponential regulator and, more importantly, the two-loop quark jet function for the EEC in the back-to-back limit (Eq. 4.16). That jet function is the stated last missing ingredient for N3LL resummation, and it is plausibly correct: the paper derives it from Mellin moments of the TMDFFs, and the resulting endpoint cross section passes several nontrivial checks, including the leading-transcendental comparison with N=4 SYM and the momentum-conservation sum rule. Those checks are real evidence, not cosmetic consistency conditions.\n\nThe technical execution looks solid. IBP reduction, canonical differential equations, and HPLs are all standard tools, and the paper's way of applying the exponential regulator to the total emitted momentum is clean enough that it makes a future N3LO calculation plausible. Reproducing the known NNLO TMDPDF at O(epsilon^0) is a good sanity check. The scale-dependent structure reconstructed from RGEs matches the computed scale-independent terms, so the internal logic is coherent.\n\nThe soft spots are in proportion. The paper is honest about the q->q TMDFF discrepancy with Ref. [40], but it does not resolve it; it attributes the difference to the universal TMD soft function, and the checks provided do not uniquely pin down the color-mixed, subleading-transcendental constant where the discrepancy sits. The zero-bin subtraction in Eq. (2.27) is stated rather than derived, and since it is load-bearing for the TMDFF endpoint terms, a referee should push for a derivation or at least a more explicit demonstration that the exponential regulator's zero-bin contribution equals the TMD soft function exactly, not just in the leading divergent part. The lack of visible O(epsilon^2) expressions (they are in an attached file) and the absence of ancillary code make the audit trail thinner than ideal; this is addressable by asking the authors to include the file or provide additional numerical checks. These are addressable concerns, not fatal flaws.\n\nWho is this for? Anyone working on TMD factorization, EEC resummation, or precision QCD at the two-loop level and beyond. The paper deserves peer review rather than desk rejection, and I would cite it for the EEC jet function even while keeping a watchful eye on the TMDFF discrepancy.","headline":"A careful NNLO calculation that delivers the last two-loop ingredient for EEC resummation, with an honest discrepancy in the TMDFF sector that needs referee scrutiny.","tokens_in":32625,"tokens_out":1374,"would_cite":true,"duration_ms":17728,"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":"This paper establishes that the exponential regulator treats quark transverse-momentum-dependent PDFs and fragmentation functions consistently at NNLO, and extracts from them the two-loop jet function that completes N3LL resummation of…","keywords":["transverse momentum dependent PDF","transverse momentum dependent fragmentation function","rapidity divergences","exponential regulator","energy-energy correlator","N3LL resummation","soft-collinear effective theory","NNLO matching coefficients"],"falsifier":"A direct, independent calculation of the two-loop coefficient $c_2^J$, for example by computing the EEC in the back-to-back limit from full matrix elements without invoking TMD factorization, would settle the claim: if the extracted $\\delta(1-z)$ constant disagrees with Eq. (4.19), then the TMDFF result and the exponential-regulator framework are wrong.","tokens_in":31489,"feed_emoji":"⚛️","tokens_out":8944,"duration_ms":92972,"temperature":0.7,"pith_summary":"This paper establishes that the exponential regulator, an alternative scheme for taming rapidity divergences in transverse-momentum-dependent factorization, works consistently for the quark transverse-momentum-dependent parton distribution function (TMDPDF) and fragmentation function (TMDFF) at next-to-next-to-leading order (NNLO). Using it, the authors compute the full matching coefficients for quark TMDPDFs and TMDFFs to order $\\alpha_s^2$, including the $\\mathcal{O}(\\epsilon^2)$ terms in dimensional regularization that a future N3LO calculation needs. The calculation also produces the two-loop quark jet function for the energy-energy correlator (EEC) in the back-to-back limit, which had been the only missing ingredient for resumming that observable at N3LL accuracy. If correct, the framework makes those N3LL predictions possible and settles a small discrepancy with an earlier NNLO fragmentation-function result.","feed_headline":"Two-loop jet function completes N3LL energy-energy correlator","feed_subtitle":"It also settles a disputed term in quark fragmentation, unlocking N3LL predictions for energy-energy correlators.","key_machinery":"The carrying mechanism is the exponential regulator, defined in momentum space by inserting $\\exp(-b_0\\tau k^0)$ in every phase-space integral and taking $\\tau\\to 0$ after integration; at the operator level it shifts the field points in the TMD definitions by $(-ib_0\\tau,-ib_0\\tau,b_\\perp)$. Because the regulator acts on the total momentum of extra emissions, it factorizes multi-emission phase space and preserves the structure of cut propagators, so the double-real contributions can be reduced with integration-by-parts identities and solved with canonical differential equations. The remaining rapidity singularities are expanded, via identities such as Eq. (3.30), directly into delta functions and plus-distributions. The zero-bin subtraction, Eq. (2.27), which subtracts exactly the TMD soft function from the collinear and fragmentation sectors, removes the double counting, and the final two-loop jet function emerges as the Mellin moment $\\int_0^1 dx\\,x\\,C_{iq}(x,b_\\perp/x,\\mu,\\nu)$ of the TMDFF matching coefficients.","core_discovery":"The paper's central claim is that the exponential regulator is a consistent, simplifying regulator for rapidity divergences in both the soft and collinear sectors, and that it delivers the first $\\mathcal{O}(\\epsilon^2)$ bare NNLO results for the quark TMDPDF and TMDFF matching coefficients. The regulator multiplies each real-emission phase-space measure by $\\exp(-b_0\\tau k^0)$ and takes $\\tau\\to 0$ after integration, turning rapidity singularities into plus-distributions and delta functions at the integrand level. With a zero-bin subtraction that identifies the overlap with the TMD soft function, the TMDPDF coefficients fully agree with earlier NNLO calculations, but the TMDFF coefficients differ from the earlier fragmentation calculation in the coefficient of a $C_AC_F\\pi^4\\delta(1-z)$ term. The authors validate their TMDFF result by using it to compute the two-loop quark jet function for the EEC in the back-to-back limit; the new constant $c_2^J$ is the last missing ingredient for N3LL resummation of that observable, and the resulting endpoint distribution passes the leading-transcendentality check and the momentum-conservation sum rule.","pith_inferences":["The same regulator plus the $\\mathcal{O}(\\epsilon^2)$ expressions should carry directly to the gluon TMDPDF and TMDFF case, where other regulators are known to be harder; the paper leaves this to future work.","Because the regulator is defined at the operator level, its path-shifted TMD definitions could be connected to non-perturbative determinations of the same objects, for instance through Euclidean or lattice methods, a step the paper does not take.","A practical test of the claim is to compute a full N3LL+NNLO back-to-back EEC spectrum using $c_2^J$ and compare it to $e^+e^-$ event-shape data; agreement would independently confirm the disputed TMDFF term.","The resolution implies that $\\delta(1-z)$ constants in TMDFFs are fixed by universal soft physics; recomputing the same Mellin moment with an independent rapidity regulator would make that point explicit."],"forward_implications":["The two-loop jet function constant $c_2^J$ completes the perturbative input for N3LL resummation of the energy-energy correlator in the back-to-back limit.","The bare NNLO TMDPDF and TMDFF expressions through $\\mathcal{O}(\\epsilon^2)$ provide the missing perturbative input for extending the matching to N3LO.","The exponential regulator is a consistent rapidity regulator in the collinear and fragmentation sectors, not just for soft functions, so the same simplified machinery can be reused for other transverse-momentum-dependent observables.","The disputed $C_AC_F\\pi^4\\delta(1-z)$ term in the quark TMDFF is resolved: the EEC jet function built from these TMDFFs passes the leading-transcendentality and sum-rule checks, supporting the new fragmentation result.","The factorization setup is universal and applies to processes such as Drell-Yan, SIDIS, and $e^+e^-$ jet production at small transverse momentum."],"supporting_citations":[{"why":"It introduces the exponential regulator that is the paper's central tool.","marker":"[43]"},{"why":"It demonstrates the same regulator on the TMD soft function at N3LO and supplies the rapidity anomalous dimensions used here.","marker":"[50]"},{"why":"It provides the prior NNLO quark-to-quark TMDPDF result that the new calculation must and does reproduce.","marker":"[38]"},{"why":"It gives the earlier full NNLO TMDPDF calculation used as the main comparison for the distribution-sector results.","marker":"[39]"},{"why":"It gives the earlier NNLO TMDPDF and TMDFF calculation whose fragmentation-function result the paper finds does not agree in one delta-function term.","marker":"[40]"},{"why":"It provides the all-order factorization for the EEC in the back-to-back limit and the hard and soft functions that the new jet function completes.","marker":"[35]"},{"why":"It supplies the analytical NLO EEC cross section used to verify the plus-distribution endpoint terms and the momentum-conservation sum rule.","marker":"[82]"},{"why":"It supplies the supersymmetric-theory endpoint result used to check the leading-transcendental part of the new constant term.","marker":"[84]"},{"why":"It gives the previous extraction of the delta-function endpoint term via the energy-conservation sum rule, against which Eq. (4.19) is checked.","marker":"[83]"}],"fun_headline_variants":["Exponential regulator cracks NNLO quark TMDs to full epsilon order","Two-loop EEC jet function completed by exponential regulator","Quark fragmentation NNLO term settled, EEC resummation unlocked","Exponential regulator simplifies NNLO quark TMDs, completes EEC","New regulator yields NNLO quark TMDs, resolves fragmentation dispute"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the exponential regulator and the zero-bin subtraction—which removes the overlap between the collinear and fragmentation sectors and the soft sector by subtracting exactly the TMD soft function—are together a consistent scheme in those sectors at two loops.","fun_headline_variants_meta":{"raw":{"variants":["Exponential regulator cracks NNLO quark TMDs to full epsilon order","Two-loop EEC jet function completed by exponential regulator","Quark fragmentation NNLO term settled, EEC resummation unlocked","Exponential regulator simplifies NNLO quark TMDs, completes EEC","New regulator yields NNLO quark TMDs, resolves fragmentation dispute"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000927,"raw_usage":{"total_tokens":3970,"prompt_tokens":945,"completion_tokens":3025,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":561,"completion_tokens_details":{"reasoning_tokens":2934}},"tokens_in":561,"tokens_out":3025,"duration_ms":20163,"temperature":1.0,"reasoning_tokens":2934,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:00:25.691291+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct, independent calculation of the two-loop coefficient $c_2^J$, for example by computing the EEC in the back-to-back limit from full matrix elements without invoking TMD factorization, would settle the claim: if the extracted $\\delta(1-z)$ constant disagrees with Eq. (4.19), then the TMDFF result and the exponential-regulator framework are wrong.","supporting_citations":[],"review_version":1}