{"id":"53f33a7a-5d97-4305-8542-a62742a0d2d6","arxiv_id":"2505.13395","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The PanScales recipe combines analytically known ingredients to produce parton showers that are demonstrably accurate at NNLL order for global event shapes and soft-sensitive non-global observables in e+e- collisions.","lead":"This paper summarizes the recipe the PanScales team used to build the first parton showers with certified next-to-next-to-leading-logarithmic accuracy. It matters because these improved simulations could eventually change how collider data are interpreted at the LHC and future colliders.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claim that the listed ingredients yield NNLL for 'any global event shape' is conditional on an unproven exclusion: the paper itself says triple-collinear dynamics is needed for general NNLL accuracy, yet no argument shows global event shapes lie outside that class.","rationale":"The paper is a proceedings, not the primary derivation; the actual demonstrations are in Refs. 16 and 1. The strongest claim is broad ('any global event shape'), while the text provides only a list of ingredients with pointers to the literature. The reader correctly identified the completeness assumption as the weakest point, and I agree. The internal text strengthens this concern: Section 4 explicitly says triple-collinear dynamics is necessary for general NNLL accuracy, so the central claim is only tenable if global event shapes are precisely the subclass not requiring it. That exclusion is not demonstrated here. The additional observation about averaging B2(z) makes the concern concrete rather than a generic completeness worry. I do not treat disagreement with consensus, nor the absence of a full derivation in a proceedings, as a refutation; the cited PRLs provide genuine external support, and the underlying result may well be correct. Therefore the appropriate verdict is unchanged: the claim should remain conditional until the global-event-shape exclusion is shown, either in this text or by direct reference to a derivation in Refs. 1/16.","tokens_in":4419,"tokens_out":8582,"duration_ms":93600,"concrete_test":"Take a global event shape not shown in Fig. 1b but with a known analytic NNLL resummation, e.g. C-parameter or heavy jet mass. Implement the recipe exactly as stated (B2 averaged over z, ΔK1, K2, drift corrections) and compare the resulting shower's NNLL coefficient, i.e. the α_s^2 (α_s L)^n contribution with α_s(MZ)=0.118, to the analytic result of Banfi–Dreyer–Monni [27] or an equivalent reference. Exact agreement for the full L-range would resolve the concern; any discrepancy at the NNLL order would confirm that either triple-collinear dynamics or the shape-dependent part of B2 enters global event shapes.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 4 is that the enumerated ingredients give the first NNLL-accurate shower 'for any global event shape'. Immediately afterward, the paper states that 'the attainment of the correct triple-collinear dynamics' is 'necessary for general NNLL accuracy'. The paper therefore draws a line between global event shapes and a broader 'general' class, but it never proves that global event shapes receive no NNLL-level triple-collinear contribution. The global-shape recipe in Section 3 consists of: NLO matching for the hard region, ΔK1 and the double-soft acceptance for the soft region, K2 plus drift corrections for soft-collinear emissions, and 'the average NLO correction' B2(z) for hard-collinear emissions. Completeness of this list is load-bearing. A sharper untested point is the averaging of B2(z): if an event shape's NNLL coefficient weights z non-trivially, replacing B2(z) by an unweighted average may miss a shape-dependent contribution. If either the triple-collinear exclusion or the B2-averaging step fails, the strong 'any global event shape' claim is not supported by this proceedings.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings paper summarizes the PanScales collaboration's recipe for constructing parton showers with claimed next-to-next-to-leading-logarithmic (NNLL) accuracy in e+e- collisions. The recipe consists of four ingredients: NLO matching for the hard region, double-soft matrix-element corrections together with a drift correction Delta K1 for the soft region, the two-loop coupling scheme with K2 and drift corrections for the soft-collinear region, and an averaged NLO correction B2(z) for the hard-collinear region. The paper claims that this combination yields the first parton shower with NNLL accuracy for any global event shape and for non-global observables primarily sensitive to soft emissions, and it shows a comparison with ALEPH data for thrust and the Durham jet resolution y23. The derivations of the individual ingredients and the numerical logarithmic-accuracy tests are delegated to the collaboration's earlier papers, in particular Refs. 1, 16, and 25.","tokens_in":4653,"tokens_out":5015,"duration_ms":46613,"significance":"If the central claim is correct, this is an important milestone: the first parton showers with demonstrable NNLL accuracy for broad classes of e+e- observables, connecting analytic resummation with general-purpose Monte Carlo simulations. The paper is useful as a high-level statement of the PanScales strategy and is honest about an important limitation, namely that triple-collinear dynamics is still needed for 'general NNLL accuracy.' However, this proceedings does not itself demonstrate the central claim: the technical derivations and accuracy tests are not included, and the data comparison in Fig. 1b is an illustration of improved agreement rather than a quantitative log-accuracy test. The significance for the field is therefore real but conditional on the cited companion papers.","major_comments":[{"comment":"The central claim that the combination of ingredients yields the first NNLL-accurate parton shower for any global event shape is stated without demonstration in this manuscript. The derivations of K1, Delta K1, the double-soft acceptance probability, B2, and the logarithmic-accuracy tests are all delegated to Refs. 1, 16, and 25. Figure 1b shows improved agreement with ALEPH data but is not a quantitative test of logarithmic accuracy. As written, the paper does not provide evidence for the word 'demonstrably' in the abstract. Please present a brief accuracy test, such as a comparison with analytic NNLL predictions for logarithmic moments, or explicitly reframe the claim as reporting results established in the cited papers.","section":"Section 4 (Concluding remarks)"},{"comment":"The manuscript asserts that global event shapes already achieve NNLL accuracy despite the statement that 'the attainment of the correct triple-collinear dynamics' is 'necessary for general NNLL accuracy.' No argument is given to show that global event shapes do not receive NNLL-level triple-collinear contributions. This distinction is load-bearing: if a triple-collinear term contributes to the NNLL coefficient of, say, thrust or C-parameter, the central claim fails. Please provide the explicit argument or cite the precise result in Refs. 1 or 16 that establishes the exclusion for global event shapes.","section":"Section 4 (Concluding remarks) and Section 3 (hard-collinear region)"},{"comment":"The 'average NLO correction B2(z)' is not defined: it is unclear whether the z-integral is weighted by the observable's measurement function or is an unweighted average over the full z range. For a general global event shape, the hard-collinear contribution to the NNLL coefficient can depend on how the observable responds to the splitting fraction z, so an unweighted average may not be shape-independent. Please state the definition of the average and justify its universality, or restrict the claim to observables for which the averaging is valid.","section":"Section 3 (NNLL corrections to global event shapes from the hard-collinear region)"},{"comment":"Several central ingredients of the recipe are introduced only verbally: the drift corrections Delta K1(y), Delta ln k_t,sc, and Delta ln z_sc appear in the text without definitions or explicit formulas. Since these corrections are part of the claimed NNLL construction, the reader cannot verify the recipe from this manuscript alone. Please provide the definitions or give the specific equation numbers in Refs. 1 and 16 where they are defined.","section":"Section 3 (soft and soft-collinear regions)"}],"minor_comments":[{"comment":"The phrase 'PanScalescollaboration' is missing a space and should read 'PanScales collaboration'.","section":"Section 1"},{"comment":"The typesetting of Eq. (1) is inconsistent: 'alpha_MC^s' and 'alpha_s(k_t)' use different spacing and subscript/superscript conventions. Please unify the notation.","section":"Equation (1)"},{"comment":"References 16 and 28 have incomplete article numbers, reading '16' and '8' respectively; please complete the bibliographic information.","section":"References"},{"comment":"The legend labels 'PGsdf 0 PG0 PG1/2' are unclear; please identify which curve corresponds to which shower variant in a way that is legible in print.","section":"Figure 1b"},{"comment":"The citation order in the sentence listing NLL showers, [9,10,11,13,12,14,15], is not sequential; please reorder the references.","section":"Section 1 (citations)"},{"comment":"The abstract claims 'first demonstrably NNLL-accurate parton showers' without specifying the collision type, while Section 4 states the claim for e+e- collisions. Please harmonize the scope of the claim.","section":"Abstract and Section 4"}],"recommendation":"major_revision","confidential_remarks":"This is a proceedings-style summary whose main claim rests on the collaboration's own recent papers. The heavy reliance on Refs. 1, 16, and 25 is not by itself inappropriate for a proceedings, but the word 'demonstrably' in the abstract is stronger than what this manuscript alone supports. The triple-collinear exclusion for global event shapes is the point most worth probing in review: if the authors cannot supply a concrete argument or citation for why global observables are insensitive to triple-collinear NNLL terms, the central claim should be softened. I would not reject the paper, as the underlying program appears serious and the data comparison is encouraging, but the manuscript should be revised to make the status of the proof clear."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a good summary of an important result, but it is a summary, not a primary source. The first NNLL-accurate PanScales showers were published in the cited PRLs, and this text adds no new derivation, data, or algorithm. If you want the proof, you go to Refs. 1 and 16.\n\nWhat the paper does well: it collects the recipe in one place and explains it clearly. The Lund-plane picture is a nice organizing device, and the list of ingredients—NLO matching for the hard region, double-soft acceptance, ΔK1, K2, the drift corrections, and the averaged B2—is presented in a way that a non-specialist can follow. The comparison to ALEPH data in Fig. 1b is also useful: the NNLL shower visibly improves on the NLL one, which is credible evidence that the construction is behaving as intended.\n\nThe soft spots are real but not fatal. Section 4 claims NNLL accuracy for \"any global event shape\" without proving that global event shapes are outside the class of observables that need triple-collinear dynamics. The paper itself says triple-collinear dynamics is necessary for general NNLL accuracy, so the line between \"global\" and \"general\" is doing a lot of work, and no argument for the exclusion is given here. Similarly, the averaging of B2(z) over z is stated without justification for why an unweighted average suffices for any global shape. These are legitimate gaps, and the stress-test note lands on the right spot. That said, this is a proceedings; the missing derivations are in the cited PRLs, so the gaps are limitations of the format rather than red flags about the underlying physics.\n\nThe citation pattern is appropriate: the collaboration cites its own published papers for the technical steps, which is exactly what a proceedings should do. There is no external verification in this text, but the PRLs provide the support.\n\nWho gains from this: anyone who wants a quick map of the PanScales strategy without digging through three PRLs, or someone teaching parton showers who needs a compact overview. It is not a new result and should not be cited as one. I would send it to peer review, but with a request to qualify the abstract and Section 4 statements—add a sentence saying the proof is in Refs. 1 and 16, and soften \"any global event shape\" to something like \"the classes of global event shapes studied in Refs. 1 and 16.\" The underlying work is solid, so this is a correction in presentation, not a rejection.","headline":"A clean, readable proceedings summary of a genuine PanScales milestone, but the NNLL claim is stated more boldly than the text can support; all load-bearing derivations live in the cited PRLs.","tokens_in":5181,"tokens_out":2556,"would_cite":false,"duration_ms":25812,"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 paper claims a recipe that produces the first parton showers with demonstrable next-to-next-to-leading-logarithmic (NNLL) accuracy for global event shapes and for soft-sensitive non-global observables in e+e− collisions.","keywords":["parton showers","NNLL resummation","global event shapes","non-global observables","Lund plane","e+e− collisions","QCD","Monte Carlo event generators"],"falsifier":"A direct check would compare this shower's thrust distribution to a complete analytic next-to-next-to-leading-log resummation; a residual difference at relative order $\\alpha_s^2$ with logarithmic enhancement beyond the included terms would falsify the completeness claim. More narrowly, a fixed-order calculation demonstrating that triple-collinear splitting contributes to the NNLL thrust coefficient would directly refute the assertion that these observables are NNLL without triple-collinear dynamics.","tokens_in":4205,"feed_emoji":"⚛️","tokens_out":13421,"duration_ms":124574,"temperature":0.7,"pith_summary":"Parton showers are the default tool for turning QCD predictions into fully exclusive collider-event descriptions, but the showers used at the LHC have long been limited to leading or next-to-leading logarithmic accuracy. This paper's central claim is that a specific combination of corrections—NLO matching, a double-soft emission correction, the universal constants $K_1$ and $K_2$, and drift-compensation factors—yields the first parton showers with next-to-next-to-leading-logarithmic (NNLL) accuracy for any global event shape in e+e− collisions, and for non-global observables that are primarily sensitive to soft emissions. If the claim holds, these showers close much of the formal-accuracy gap between analytic resummation and Monte Carlo event generators, giving collider phenomenology an exclusive event generator with NNLL logarithmic accuracy. The paper also states that genuine triple-collinear dynamics, while not needed for these observables, remains necessary for general NNLL accuracy.","feed_headline":"First parton showers hit NNLL accuracy for event shapes","feed_subtitle":"A recipe of constants and recoil-drift fixes gives e+e− event-shape predictions NNLL precision.","key_machinery":"The Lund plane—a map of radiative phase space with $\\ln k_t$ and rapidity $y$ as coordinates—is the central bookkeeping device: it sorts emissions into hard-collinear, soft large-angle, and soft-collinear regions and tracks how later branchings drift earlier emissions. The other load-bearing object is the Monte Carlo scheme for the strong coupling, $\\alpha_s^{\\mathrm{MC}}(k_t)=\\alpha_s(k_t)[1+(\\alpha_s/2\\pi)K_1]$, which encodes the universal single-emission constant. Around these, the recipe adds ratio corrections from the exact double-soft matrix element, the NNLO constant $K_2$, the hard-collinear average $B_2(z)$, and drift factors $\\Delta K_1$, $\\Delta\\ln k_t$, and $\\Delta\\ln z$ that compensate for coordinate shifts introduced by the shower's momentum mapping.","core_discovery":"The discovery is a constructive recipe: start from an NLL-accurate dipole shower, then add (i) NLO matching to capture the $\\alpha_s(\\alpha_s L)^n$ terms, (ii) a correction to the double-soft emission rate when two soft partons are close in the Lund plane, (iii) the universal constants $K_1$ and $K_2$ in a properly defined Monte Carlo $\\alpha_s$ scheme with three-loop running, (iv) drift-compensation factors $\\Delta K_1(y)$, $\\Delta\\ln k_t$, and $\\Delta\\ln z$ that correct for the way the shower's kinematic map shifts emissions in rapidity and transverse momentum, and (v) the average hard-collinear correction $B_2(z)$. The paper asserts that this combination is sufficient to make the shower NNLL-accurate for any global event shape, and for non-global observables mainly sensitive to soft emissions; it further reports improved agreement with e+e− event-shape data once these corrections are included.","pith_inferences":["One could view the drift-compensation factors as a general design rule: any shower whose recoil strategy shifts an emission's resolved coordinates must locally recompute the scheme-dependent constants, a principle likely to transfer to hadron-collider showers.","A testable extension is to apply the same recipe to further global event shapes, such as C-parameter or heavy-jet mass, and compare against analytic NNLL resummations; flat ratios would corroborate the completeness assumption.","If the completeness assumption is later found to fail for some global observable, the likely culprit is a triple-collinear term, making the explicit omission of triple-collinear dynamics the clearest target for scrutiny."],"forward_implications":["If the recipe is correct, e+e− event-shape predictions from these showers carry NNLL logarithmic accuracy, meaning their logarithmic uncertainty is of order $\\alpha_s^2(\\alpha_s L)^n$ rather than $\\alpha_s(\\alpha_s L)^n$.","For non-global observables dominated by soft emissions, the shower achieves the same accuracy as state-of-the-art analytic resummations, not just a formal claim but a demonstrable one.","The corrections improve agreement with LEP event-shape data for thrust and the Durham jet-resolution variable $y_{23}$, as shown in the paper's comparison plots.","The paper leaves correct triple-collinear dynamics and extensions to incoming partons and multiple hard jets as the next milestones, implying the present accuracy does not yet cover all hadron-collider processes."],"supporting_citations":[{"why":"The companion analysis that demonstrates the NNLL accuracy of the showers and underpins the central claim.","marker":"1"},{"why":"Supplies the exact double-soft matrix element ratio used to correct the shower's soft-emission rate.","marker":"16"},{"why":"Defines the Monte Carlo scheme for $\\alpha_s$ and the universal constant $K_1$ that carries the single-logarithmic corrections.","marker":"17"},{"why":"Provides the NNLO inclusive correction $K_2$ required for soft-collinear emissions.","marker":"18"},{"why":"Gives the average NLO hard-collinear correction $B_2$ used in the hard-collinear region.","marker":"19"},{"why":"Supplies the calculation of $B_2(z)$ and its ingredients for the hard-collinear corrections.","marker":"20"},{"why":"Shows how NLO matching contributes the $\\alpha_s(\\alpha_s L)^n$ terms without double-counting the hardest emission.","marker":"25"},{"why":"Introduces the Lund plane, the phase-space map on which the shower corrections and drift compensation are defined.","marker":"22"},{"why":"Provides the e+e− event-shape data the shower predictions are compared against.","marker":"29"}],"fun_headline_variants":["PanScales recipe yields first NNLL-accurate parton showers","First NNLL parton showers: the PanScales construction details","How to build NNLL parton showers: PanScales' recipe","PanScales' five-step recipe for NNLL parton showers","From NLL to NNLL: PanScales' new shower recipe"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that, for global event shapes, the listed corrections are complete at this order of log accuracy, so genuine three-parton collinear splitting dynamics is not needed for those observables.","fun_headline_variants_meta":{"raw":{"variants":["PanScales recipe yields first NNLL-accurate parton showers","First NNLL parton showers: the PanScales construction details","How to build NNLL parton showers: PanScales' recipe","PanScales' five-step recipe for NNLL parton showers","From NLL to NNLL: PanScales' new shower recipe"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000244,"raw_usage":{"total_tokens":1456,"prompt_tokens":795,"completion_tokens":661,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":411,"completion_tokens_details":{"reasoning_tokens":567}},"tokens_in":411,"tokens_out":661,"duration_ms":6302,"temperature":1.0,"reasoning_tokens":567,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:14:43.893647+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct check would compare this shower's thrust distribution to a complete analytic next-to-next-to-leading-log resummation; a residual difference at relative order $\\alpha_s^2$ with logarithmic enhancement beyond the included terms would falsify the completeness claim. More narrowly, a fixed-order calculation demonstrating that triple-collinear splitting contributes to the NNLL thrust coefficient would directly refute the assertion that these observables are NNLL without triple-collinear dynamics.","supporting_citations":[{"cited_title":"van Beekveld, M","cited_arxiv_id":null,"evidence_quote":"The companion analysis that demonstrates the NNLL accuracy of the showers and underpins the central claim."},{"cited_title":"Ferrario Ravasio, K","cited_arxiv_id":null,"evidence_quote":"Supplies the exact double-soft matrix element ratio used to correct the shower's soft-emission rate."},{"cited_title":"Catani, B","cited_arxiv_id":null,"evidence_quote":"Defines the Monte Carlo scheme for $\\alpha_s$ and the universal constant $K_1$ that carries the single-logarithmic corrections."},{"cited_title":"Banfi, B","cited_arxiv_id":null,"evidence_quote":"Provides the NNLO inclusive correction $K_2$ required for soft-collinear emissions."},{"cited_title":"Dasgupta and B","cited_arxiv_id":null,"evidence_quote":"Gives the average NLO hard-collinear correction $B_2$ used in the hard-collinear region."},{"cited_title":"van Beekveld, M","cited_arxiv_id":null,"evidence_quote":"Supplies the calculation of $B_2(z)$ and its ingredients for the hard-collinear corrections."},{"cited_title":"Hamilton, A","cited_arxiv_id":null,"evidence_quote":"Shows how NLO matching contributes the $\\alpha_s(\\alpha_s L)^n$ terms without double-counting the hardest emission."},{"cited_title":"Andersson, G","cited_arxiv_id":null,"evidence_quote":"Introduces the Lund plane, the phase-space map on which the shower corrections and drift compensation are defined."},{"cited_title":"Heisteret al.[ALEPH], Eur","cited_arxiv_id":null,"evidence_quote":"Provides the e+e− event-shape data the shower predictions are compared against."}],"review_version":1}