{"id":"cf17f3c2-1dc7-4fd4-a7aa-1a95958d780a","arxiv_id":"1908.09848","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A KZB-type differential equation and Picard iteration yield the all-order alpha-prime expansion of one-loop open-string integrals in terms of iterated Eisenstein integrals.","lead":"The paper presents a new differential-equation method to compute the full alpha-prime expansion of genus-one open-string integrals, with coefficients organized in elliptic multiple zeta values. The approach connects string-amplitude computations to modern number theory and should enable systematic higher-order results in string perturbation theory.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (8) is the linchpin: the paper defers its derivation and only verifies n=2,3; if the boundary terms in the integration by parts do not vanish at n≥4, the Picard solution and the main expansion (14) fail.","rationale":"The reader's conditional verdict is appropriate. The paper's stated objective is an all-order α'-expansion of one-loop open-string integrals, and its main formula (14) is derived from the differential equation (8) via Picard iteration. The derivation of (8) is explicitly deferred to a companion paper and depends on vanishing boundary terms that are only demonstrated for n=2 and n=3. This is exactly the point where the central claim could fail at higher multiplicity, because boundary terms are not fixed by the local differential identities (5) and (9); they depend on the global integration cycle. The paper itself signals this by saying the boundary terms vanish, but provides no proof in the letter. The Tsunogai-representation conjecture is real but secondary: it concerns the minimal eMZV form of the coefficients, not the validity of the expansion in iterated Eisenstein integrals. My proposed check—an explicit four-point derivation of (8) with all boundary terms evaluated—would settle the main concern. Since the reader already marked the verdict CONDITIONAL with medium confidence, and my analysis identifies the same weakest assumption, no verdict change is needed.","tokens_in":10891,"tokens_out":3668,"duration_ms":40653,"concrete_test":"Independently derive (8) at n=4. Differentiate the four-point A-cycle integral (6) with respect to τ using (5) and (9), perform integration by parts in v_2, v_3, v_4, and explicitly evaluate all boundary terms for both the planar domain 0<v_2<v_3<v_4<1 and the non-planar domain with u_j=1/2. If the boundary sums vanish identically for generic s_ij, equation (8) holds at n=4 and the main risk is retired; if they do not vanish, the central expansion (14) fails at the first unverified multiplicity. A cross-check at the same order is to compare the α'^2 terms of the four-point expansion from (14) with the direct puncture integration using known eMZV results from [6].","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (14) rests entirely on the universal differential equation (8), whose derivation is not given in this letter. The text says (8) follows from the Green-function properties (5), the mixed heat equation (9), and 'the vanishing of boundary terms ∫ dv_j ∂v_j(...)'. That last condition is the only place where the integration domain can obstruct the result, and it is verified in the letter only through the two- and three-point examples of Section 2. At n=4 and beyond, the A-cycle integrals mix under τ-derivatives, and the boundary terms involve coincident-puncture limits of Kronecker–Eisenstein series as well as, for non-planar cycles, the explicit τ-dependence of the puncture positions z_j = τ/2 + v_j. If those boundary contributions are nonzero for generic Mandelstam invariants, equation (8) would acquire an inhomogeneous or different-cycle term, so the Picard iteration (10) and the factorized form (14) would not follow. The companion paper [12] is cited for the all-multiplicity derivation, but the letter itself does not establish the needed identity. A secondary, explicitly conjectural ingredient is that r_η(ϵ_k) preserves Tsunogai's derivations; this affects the claim that the coefficients are eMZVs in minimal form, but it would not invalidate the α'-expansion as iterated Eisenstein integrals. Therefore equation (8) is the most load-bearing assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a new method for the all-order α'-expansion of one-loop open-string integrals. It introduces generating functions Z^τ_η for A-cycle integrals on the cylinder and claims that they satisfy a KZB-type differential equation (8). Solving (8) by Picard iteration, the paper expresses Z^τ_η as a sum of iterated Eisenstein integrals γ(k_1,...,k_r|τ) multiplied by matrices r_η(ε_k) acting on cusp initial values Z^{i∞}_η, which in turn are reduced to genus-zero disk integrals. The authors work out the differential operators and cusp initial values explicitly for n=2 and n=3, and argue that the structure matches Tsunogai's derivations and elliptic multiple zeta values in minimal form. The main result is equation (14), with the all-multiplicity derivation of the underlying differential equation deferred to the companion paper [12].","tokens_in":11220,"tokens_out":5127,"duration_ms":53582,"significance":"If the central differential equation (8) and the cusp reduction hold, the paper provides a conceptually new and potentially powerful method: the α'-expansion of one-loop open-string integrals is organized by iterated Eisenstein integrals with genus-zero initial data, connecting directly with the elliptic KZB associator and Tsunogai's derivations. The explicit n=2 and n=3 examples are concrete and checkable, and the letter is clearly written. The main strengths are the elegance of the proposed structure and the explicit verifications that support it. However, the letter's central claims rest on a small number of unproved or only partially verified ingredients, so the significance is conditional on those ingredients being established in the companion paper.","major_comments":[{"comment":"The universal differential equation (8) is the sole foundation for the Picard solution (10) and the main expansion (14), yet its derivation is not supplied in this letter. The text states that (8) follows from (5), (9) and \"the vanishing of boundary terms ∫ dv_j ∂v_j(...)\", and the only checks shown are the n=2 and n=3 cases of Section 2. For n≥4 the τ-derivatives mix permutations and, for non-planar cycles, the punctures at z_j = τ/2+v_j carry explicit τ-dependence; if any boundary contribution survives, equation (8) would acquire extra terms and the factorized form (14) would not hold. I ask the authors to either include a proof or a detailed outline of the boundary-term cancellation at generic multiplicity, or a nontrivial n=4 example with planar and non-planar cycles, or to state explicitly in the letter that (8) is a theorem proved in [12] and state its precise hypotheses. As written, the central claim depends on an unstated assumption.","section":"§2B, Eq. (8)"},{"comment":"The statement that the coefficients in (14) are elliptic multiple zeta values in minimal form rests on the assertion that r_η(ε_k) \"should preserve the commutation relations of Tsunogai's derivations\" and the immediately following sentence that they \"are believed to furnish matrix representations\". This is a conjecture, not a proven property. The n=3 representation (22) checks some relations, but no all-multiplicity proof is given. Because the abstract and conclusions present the minimal eMZV statement as a result, the authors should either prove the representation property or explicitly mark it as a conjecture and adjust the claims about minimality accordingly.","section":"§2C, paragraph after Eq. (14)"},{"comment":"The reduction of cusp initial values to disk integrals is another essential input of (14), but the general n-point statement is deferred to [12]; equations (31)–(37) give only n=2,3 and rely on contour-deformation factors such as the \"2i sin(πs12/2)\" in (31), whose derivation is also deferred. Since these cusp values are the seed data for the Picard iteration, the same completeness concern applies as for (8). Please provide the general degeneration formula or a precise pointer to the statement in [12], and ideally one higher-n example so the reader can verify the form of the kinematic limits (28).","section":"§3A, Eqs. (28) and (31)-(37)"}],"minor_comments":[{"comment":"The notation r_η(ε_{kr}...ε_{k1})_{B C} is not fully defined: please clarify that the subscript denotes the matrix element with row and column labelled by the permutations B and C, and specify the ordering convention for the matrix product.","section":"Eq. (14)"},{"comment":"The non-planar A-cycle domain is only described in words; please spell out the substitution z_j = τ/2 + v_j with v_j ∈ (0,1) and the ordering of the v_j for a generic non-planar permutation, so that the integrals in (6) are unambiguous.","section":"Eq. (6) and Fig. 2"},{"comment":"The notation for the permutation map in (21), \"s12↔s13, η2↔η3\", is clear in context but the arrow notation is nonstandard; consider writing the full map of Mandelstam variables and η's explicitly.","section":"Eq. (21)"},{"comment":"The appearance of the factor 2i sin(πs12/2) is stated to stem from contour deformations detailed in [12]; a one-line explanation or an equation reference inside [12] would help the reader follow the derivation on the page.","section":"§3B, Eq. (31)"},{"comment":"The convention G_0 = -1 is introduced in the text, but it may be worth noting explicitly that this is a normalization of the Weierstrass expansion and not a standard Riemann zeta value, to avoid confusion with G_2 and G_4.","section":"§2B, Eq. (12)"}],"recommendation":"major_revision","confidential_remarks":"The central technical content is deferred to the companion paper [12] by the same authors. The editor may wish to confirm that [12] indeed contains the all-multiplicity proof of (8) and the general cusp reduction, and that the representation-theoretic statement about r_η(ε_k) is proved there rather than only conjectured. The phrase \"believed to furnish matrix representations\" should be made transparent in the abstract or conclusions if it remains a conjecture."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper gives a new way to compute the alpha-prime expansion of genus-one open-string integrals. Instead of direct integration, the authors write generating functions for the integrals, derive a KZB-type differential equation, and solve via Picard iteration. The result is a clean formula (14) expressing the integrals as iterated Eisenstein integrals acting on genus-zero (disk) initial data. The explicit two- and three-point examples check out, and the connection to Tsunogai derivations is genuinely suggestive. That is real progress and goes beyond earlier work [6-8].\n\nWhere it is soft: the linchpin is equation (8), the differential equation for the generating functions. The letter does not derive it; it says it follows from Green function properties, the mixed heat equation, and the vanishing of boundary terms ∫ dv_j ∂v_j(...). That last condition is precisely where the cylinder domain can bite, and it is only verified for n=2,3. The stress-test note is right to identify this as the load-bearing assumption. If boundary terms are nonzero at n≥4, the Picard solution and (14) fall apart. The companion paper is cited for the all-multiplicity proof, but this letter alone does not establish it. A referee would need to check [12] carefully.\n\nSecondary, and explicitly conjectural: r_η(ε_k) are said to preserve Tsunogai's derivations, which is needed for the coefficients to be eMZVs in minimal form. That is marked as 'believed'. The alpha-prime expansion as iterated Eisenstein integrals does not depend on it, so this is a minor caveat.\n\nAlso note the self-citation is reasonable here: [12] is the detailed paper with the derivations, not hidden data. The initial-value reduction to disk integrals is concrete and checkable.\n\nBottom line: if (8) holds, (14) is a substantial result. The letter is not fully self-contained, but for a letter that is common. I would take the claim seriously and would send it to peer review with a referee asked to verify (8) and the boundary terms, and to check the Tsunogai representation. The paper is for string-amplitude people and anyone working on elliptic associators; they will get use from it.\n\nRecommendation: accept conditional on the companion-paper verification. It deserves a serious referee.","headline":"Elegant and probably right: a KZB-type differential equation yields a compact all-order alpha-prime expansion for one-loop open-string integrals, but the key equation is asserted, not proved, in the letter.","tokens_in":11697,"tokens_out":2142,"would_cite":true,"duration_ms":22560,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11G55","11F67","81T30"],"pacs":["11.25.-w"],"model":"deepseek-v4-flash","headline":"A KZB-type differential equation reduces one-loop open-string integrals to iterated Eisenstein integrals, with tree-level cusp values as the seed data.","keywords":["open-string amplitudes","elliptic multiple zeta values","iterated Eisenstein integrals","Knizhnik-Zamolodchikov-Bernard equation","alpha-prime expansion","genus-one integrals","disk integrals","Kronecker-Eisenstein series"],"falsifier":"Compare the planar four-point A-cycle integral at the first non-leading order in $\\alpha'$ computed by direct puncture integration with the value produced by equation (14); a mismatch, or a nonzero boundary term $\\int dv_j\\,\\partial_{v_j}(\\cdots)$ in the derivation of (8) at four points, would falsify the universal differential equation and the expansion.","tokens_in":10696,"feed_emoji":"📐","tokens_out":12909,"duration_ms":115007,"temperature":0.7,"pith_summary":"One-loop open-string amplitudes live on a cylinder or Möbius strip, and their low-energy expansion in the inverse string tension $\\alpha'$ had previously been computed by direct integration over world-sheet punctures, order by order. This paper claims that a generating function of all such integrals obeys a simple Knizhnik–Zamolodchikov–Bernard-type differential equation in the torus modulus, and that Picard iteration solves it in a uniform way: every order of the $\\alpha'$-expansion is a finite combination of iterated Eisenstein integrals acting on genus-zero (tree-level) initial data. If the claim is right, the elliptic multiple zeta values that appear in these amplitudes come out automatically in their minimal form, and the expansion is governed by the same kind of first-order differential-equation structure that underlies modern Feynman-integral computations. The payoff is a uniform, all-multiplicity description of one-loop open-string expansions.","feed_headline":"One-loop open-string expansions collapse to iterated Eisenstein integrals","feed_subtitle":"One differential equation in the torus modulus yields every order from tree-level data alone.","key_machinery":"The machinery is the generating function $Z^\\tau_{\\vec{\\eta}}$ of equation (6), built from the Kronecker–Eisenstein series and the Koba–Nielsen factor, together with its universal differential equation (8): $2\\pi i\\,\\partial_\\tau Z^\\tau_{\\vec{\\eta}}(A|1,B)=\\sum_C D^\\tau_{\\vec{\\eta}}(B|C)\\,Z^\\tau_{\\vec{\\eta}}(A|1,C)$. The matrix $D^\\tau_{\\vec{\\eta}}$ is linear in the Mandelstam invariants $s_{ij}$ and therefore in $\\alpha'$, and its entire $\\tau$-dependence is carried by Weierstrass $\\wp$-functions, giving the decomposition $D^\\tau_{\\vec{\\eta}}=\\sum_k(1-k)G_k(\\tau)\\,r_{\\vec{\\eta}}(\\epsilon_k)$. Picard iteration of this first-order equation produces the iterated Eisenstein integrals $\\gamma(k_1,\\dots,k_r|\\tau)$; the initial values at the cusp $\\tau\\to i\\infty$ are fixed by known disk integrals. The action of the matrices $r_{\\vec{\\eta}}(\\epsilon_k)$ is what arranges the expansion so that elliptic multiple zeta values appear in minimal form.","core_discovery":"The central result is equation (14), an explicit all-order formula for the A-cycle integrals $Z^\\tau_{\\vec{\\eta}}(A|1,B)$ that generate one-loop open-string amplitudes: $$Z^\\tau_{\\vec{\\eta}}(A|1,B)=\\sum_{r\\ge0}\\ \\sum_{k_1,\\dots,k_r\\in\\{0,4,6,\\dots\\}} \\gamma(k_1,\\dots,k_r|\\tau)\\prod_{j=1}^r(k_j-1)\\sum_{C\\in S_{n-1}} r_{\\vec{\\eta}}(\\epsilon_{k_r}\\cdots\\epsilon_{k_1})_{BC}\\,$Z^{{i\\infty}}$_{\\vec{\\eta}}(A|1,C).$$ Here $\\gamma(k_1,\\dots,k_r|\\tau)$ are iterated Eisenstein integrals with tangential-base-point regularization, $r_{\\vec{\\eta}}(\\epsilon_k)$ are $\\tau$-independent matrix differential operators in the variables $\\eta_j$, and $Z^{i\\infty}_{\\vec{\\eta}}$ are the cusp values obtained by degenerating the cylinder to genus-zero disk integrals. The same formula covers planar, non-planar and Möbius-strip configurations, and the indices $k_j$ run only over $0,4,6,8,\\dots$ because odd and weight-two Eisenstein series drop out of the Weierstrass decomposition of the differential operator. The authors put this forward as the exact $\\alpha'$-expansion, not an approximation.","pith_inferences":["A natural next test is to push the four- and five-point cases to higher weight: if the matrices $r_{\\vec{\\eta}}(\\epsilon_k)$ obey Tsunogai's relations, the identities among iterated Eisenstein integrals should exactly match the known elliptic-multiple-zeta-value relations, which would effectively prove the representation conjecture.","Because the setup is formulated at the level of generating functions, the same differential-equation logic could be adapted to closed-string one-loop integrals, where the $\\tau$-expansion would encode modular graph forms rather than elliptic multiple zeta values.","One could implement equation (14) as a purely algebraic algorithm: precompute the finite set of matrices $r_{\\vec{\\eta}}(\\epsilon_k)$ once for a given $n$, and then each $\\alpha'$-order coefficient is a matrix product times weighted sums of iterated Eisenstein integrals, making very high orders accessible numerically.","The method suggests a strategy for higher genus: use separating and non-separating degenerations of the world-sheet as initial conditions for differential equations in the complex-structure moduli, though the paper only states this as a direction."],"forward_implications":["Every order of the $\\alpha'$-expansion of massless one-loop open-string amplitudes is obtained by evaluating finite products of the matrices $r_{\\vec{\\eta}}(\\epsilon_k)$ on tree-level initial data, so no new puncture integration is needed at higher orders.","The elliptic multiple zeta values in the coefficients are produced in minimal form, with their rational relations built in through the commutation relations of Tsunogai's derivations.","Planar and non-planar A-cycle integrals, and therefore cylinder and Möbius-strip amplitudes, are governed by the same differential operator, so the same generating functions control the cancellation of tadpole divergences.","Because the differential equation is linear in $\\alpha'$, the method gives a genus-one counterpart of the $\\varepsilon$-form of Feynman-integral differential equations, sharpening the analogy between $\\alpha'$ and the dimensional-regularization parameter $\\varepsilon$.","The initial conditions are genus-zero disk integrals, so the one-loop expansion inherits whatever structure is known at tree level, including multiple zeta values from disk integrals."],"supporting_citations":[{"why":"Supplies the derivation of the universal differential equation (8) and the all-multiplicity details, including boundary-term conditions and cusp initial values.","marker":"[12]"},{"why":"Provides the earlier direct puncture-integration method for one-loop open-string integrals whose coefficients introduce elliptic multiple zeta values.","marker":"[6]"},{"why":"Defines the elliptic Knizhnik–Zamolodchikov–Bernard associator whose differential equation has the same shape as equations (12) and (14).","marker":"[10]"},{"why":"Defines Tsunogai's derivations dual to Eisenstein series, whose commutation relations the tau-independent matrices are expected to represent.","marker":"[11]"},{"why":"Establishes the minimal form of elliptic multiple zeta values and the role of derivation algebras in counting independent values.","marker":"[9]"},{"why":"Gives the disk integrals used to evaluate the genus-zero initial data $Z^{i\\infty}_{\\vec{\\eta}}$.","marker":"[19]"},{"why":"Provides the tangential-base-point regularization that defines the iterated Eisenstein integrals $\\gamma(k_1,\\dots,k_r|\\tau)$.","marker":"[17]"}],"fun_headline_variants":["All-order one-loop string formula from a single differential equation","Iterated Eisenstein integrals yield all orders of one-loop open strings","One-loop open-string integrals: all orders from tree-level data","Exact alpha'-expansion via iterated Eisenstein integrals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the universal differential equation (8) holds at all multiplicities, which is verified only at two and three points and requires all cylinder integration-by-parts boundary terms to vanish, together with the unproven assumption that the matrices $r_{\\vec{\\eta}}(\\epsilon_k)$ preserve Tsunogai's commutation relations.","fun_headline_variants_meta":{"raw":{"variants":["All-order one-loop string formula from a single differential equation","Iterated Eisenstein integrals yield all orders of one-loop open strings","One-loop open-string integrals: all orders from tree-level data","Exact alpha'-expansion via iterated Eisenstein integrals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000921,"raw_usage":{"total_tokens":3934,"prompt_tokens":914,"completion_tokens":3020,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":530,"completion_tokens_details":{"reasoning_tokens":2950}},"tokens_in":530,"tokens_out":3020,"duration_ms":23238,"temperature":1.0,"reasoning_tokens":2950,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:59:24.981734+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the planar four-point A-cycle integral at the first non-leading order in $\\alpha'$ computed by direct puncture integration with the value produced by equation (14); a mismatch, or a nonzero boundary term $\\int dv_j\\,\\partial_{v_j}(\\cdots)$ in the derivation of (8) at four points, would falsify the universal differential equation and the expansion.","supporting_citations":[{"cited_title":"The decomposition of eMZVs into iterated Eisenstein integrals automatically incorporates all their relations over the rational numbers [9]","cited_arxiv_id":null,"evidence_quote":"Supplies the derivation of the universal differential equation (8) and the all-multiplicity details, including boundary-term conditions and cusp initial values."},{"cited_title":"The second entry Z τ ⃗ η(∗|A) specifies permuta- tions A = a1a2","cited_arxiv_id":null,"evidence_quote":"Provides the earlier direct puncture-integration method for one-loop open-string integrals whose coefficients introduce elliptic multiple zeta values."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the elliptic Knizhnik–Zamolodchikov–Bernard associator whose differential equation has the same shape as equations (12) and (14)."},{"cited_title":"Tsunogai, Publ","cited_arxiv_id":null,"evidence_quote":"Defines Tsunogai's derivations dual to Eisenstein series, whose commutation relations the tau-independent matrices are expected to represent."},{"cited_title":"Their τ -derivatives resulting from (5), (9) and integration by parts w.r.t","cited_arxiv_id":null,"evidence_quote":"Establishes the minimal form of elliptic multiple zeta values and the role of derivation algebras in counting independent values."},{"cited_title":"Levin, Compositio Mathematica 106 (1997) 267","cited_arxiv_id":null,"evidence_quote":"Provides the tangential-base-point regularization that defines the iterated Eisenstein integrals $\\gamma(k_1,\\dots,k_r|\\tau)$."}],"review_version":1}