Pith. sign in

REVIEW 2 major objections 5 minor 54 references

This paper claims that four-point conformal integrals can be evaluated by splitting them into single-singularity pieces and bootstrapping each piece, yielding four-loop results other methods cannot compute yet.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-02 06:48 UTC pith:6CDSFSBK

load-bearing objection Genuinely new four-loop conformal integral evaluations in single-valued MPLs, but the bootstrap's empirical ansatz has a stated counterexample (I_22) that should be addressed before the results are treated as established. the 2 major comments →

arxiv 2607.11645 v2 pith:6CDSFSBK submitted 2026-07-13 hep-th

Notes on the bootstrap of four-point conformal integrals

Xuhang Jiang This is my paper
classification hep-th
keywords conformal integralsbootstrap methodleading singularitiessingle-valued multiple polylogarithmsfour-point correlation functionsN=4 super-Yang-Millsf-graphsexpansion by regions
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper tries to establish that four-point conformal integrals in position space can be evaluated with a bootstrap that uses very little information from the integrand itself: leading singularities, single-valued multiple polylogarithm ansätze, and boundary values from expansion by regions. The author shows that the entire three-loop sector—fifteen integrand basis elements, organized into twelve independent analytic functions—can be determined in this way, and that some four-loop integrals that are not calculable by other known methods can also be obtained. The enabling move is a decomposition: an integral with several leading singularities is split into pieces with a single leading singularity each, and each piece is bootstrapped separately. The results are analytic expressions in single-valued polylogarithms for the four-loop integrals I_173, I_176 and I_181, together with an empirical rule for which letters enter the function space.

Core claim

The paper's central discovery is that a minimal-information bootstrap determines four-point conformal integrals that have so far resisted other methods. For the three-loop sector, fourteen of the fifteen inequivalent integrand basis elements follow directly from a symbolic integration package and the fifteenth from a Gram-determinant identity, yielding twelve independent analytic functions. At four loops the paper analyzes all 412 integrals generated by planar and non-planar f-graphs, finds 164 directly computable, and shows that several of the remaining hard cases—labelled I_173, I_176, and I_181—can be evaluated in terms of single-valued multiple polylogarithms. The enabling step is a deco

What carries the argument

The load-bearing mechanism is the splitting of a multi-leading-singularity conformal integral into single-leading-singularity pieces. A leading singularity is the residue obtained by cutting enough propagators to localize all loop vertices; for position-space conformal integrals these residues are rational functions of the cross ratios u and v. When an integral has several such residues, the paper shifts its numerator by Gram-determinant/Jacobian combinations such as x^2_18 x^2_34 − x^2_13 x^2_48 to cancel one pole at a time, producing simpler pieces. Each piece is then matched to an ansatz in single-valued multiple polylogarithms—functions built by single-valued integration with letters at

Load-bearing premise

The load-bearing premise is that the ansatz function space is complete—specifically that any new letter enters only in the last two positions and is read off from reduced-graph leading singularities—which is an observation, not a proof, so a missed letter could produce a wrong function that still matches the boundary data.

What would settle it

Compute the O(u^1, Y^4) term of the expansion-by-regions boundary for I_176^(4) and compare it with the series expansion of the bootstrapped analytic expression; any mismatch at that order would demonstrate the ansatz missed a letter or a non-last-two-entry structure.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • The complete three-loop basis of four-point conformal integrals from f-graphs is now determined analytically, organized into twelve independent functions, providing a benchmark for any future method.
  • The four-loop integrals I_173, I_176 and I_181, which are not (or not easily) accessible to direct integration, are obtained in closed form as single-valued multiple polylogarithms.
  • Decomposing an integral by its leading singularities turns a hard multi-singularity problem into several single-singularity bootstraps, which is expected to extend to other integrals whose leading singularities are rational functions of u and v.
  • For rigid polylogarithmic integrals, boundary data alone—together with the mirror constraints—can fix all ansatz coefficients, removing the need for additional magic identities in these cases.
  • The empirical rule connecting reduced-graph leading singularities to last-two-entry letters gives a practical heuristic for constructing function spaces for future conformal integrals, though it remains an observation rather than a proven principle.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the last-two-entries rule extends beyond the tested examples, it suggests a general dictionary between the cuts of reduced graphs and the function space of the parent integral; a natural test is to apply generalized boxing to a broader class of four-loop integrals with rational leading singularities and check whether the predicted letters occur at the predicted depth.
  • The mirror-constraint trick—expanding with z and z-bar interchanged and matching to the same boundaries—is likely a general technique for any finite, real-axis-smooth integral whose known expansions are of low order; it may sharpen other bootstrap setups where boundary data are scarce.
  • The decomposition of an integrand by adding and subtracting Gram-determinant numerators is reminiscent of canonical dlog-form constructions; one could try to turn it into an automated differential-equation or intersection-theory reduction for non-planar conformal integrals.
  • The paper's classification data (412 four-loop integrands, their symmetry and Gram identities) is itself a resource: it could seed searches over the same function space using the accompanying skill files designed for current AI models.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper develops a bootstrap workflow for four-point conformal integrals in position space. The workflow combines leading-singularity analysis, ansaetze in single-valued multiple polylogarithms (svMPLs), boundary data obtained from expansion by regions, and additional 'mirror constraints' coming from the no-branch-cut property on the real axis. The author applies this workflow to the three-loop sector, where all fifteen integrands are determined (one via a Gram identity), and then to selected four-loop integrals, notably I_176^(4) and I_181^(4), which are decomposed into pieces with a single leading singularity and bootstrapped separately. The paper also provides a package with AI-readable skill files and reports numerical checks with the independent integrator trillo.

Significance. If the four-loop results are correct, the paper demonstrates a meaningful extension of the conformal-integral bootstrap beyond the reach of packages such as HyperlogProcedures. The classification of 412 four-loop f-graph integrands, the explicit four-loop nonplanar master integrals, the systematic use of mirror constraints, and the successful three-loop benchmark are useful contributions. The central results, however, rest on an empirical ansatz-construction rule that the paper itself labels as 'only an observation' (App. B), and the numerical verification is comparatively coarse; these points limit the current certainty of the claimed new four-loop evaluations.

major comments (2)
  1. [Sec. 2.1.4 and App. B] The completeness of the ansatz space is load-bearing and is not established. App. B states that new letters in svMPLs appear 'only in the last two entries' and can be read off from reduced-graph leading singularities, calling this 'only an observation rather than a principle.' Section 2.1.4 then uses this rule to construct the ansatz for I_176 and I_181. Yet the same section's I_22 example shows that the letter \bar z appears in the last four entries, not just the last two, and this is confirmed by HyperlogProcedures. Thus the restricted function space can miss valid structures. The statement in Sec. 2.1.4 that the ansatz is 'justified by the success of bootstrap and the verification of numerical values' is not a completeness argument: a fit inside an incomplete space can satisfy the available boundary and still be wrong. A decisive, in-scope test would be to run the bootstrap with the p
  2. [Sec. 3.2.2 and Sec. 2.2] The numerical verification is too coarse to exclude a nearby wrong function. The paper reports only that the four-loop results are 'checked by trillo up to 10^{-4} ~ 10^{-5} relative error,' without stating the number of sample points or their location. Given that in Sec. 3.2.1 the consistent boundary choice alone fixes only 221 of 304 coefficients for I_173 and mirror constraints are needed to fix the rest, underdetermination is a real phenomenon in this workflow. A wrong function in a larger ansatz space could easily agree with the boundary data to O(u^0,Y^4) and with a 10^-4 numerical check. I recommend independent high-precision checks (at least 10^-10 at several points) and/or an exact or much-higher-order boundary comparison for I_176 and I_181.
minor comments (5)
  1. [Sec. 2.1.3, Eq. (2.27)] The split of I_176 into I_a + I_b + alpha I_c introduces an arbitrary parameter alpha. The text later suggests that alpha=0 is used, but this is stated only obliquely and the equation reference appears garbled ('(2.32)' in Sec. 2.1.4). Please state explicitly the value of alpha used in the final bootstrap, or demonstrate that the final result is independent of alpha.
  2. [Sec. 2.1.4, Eq. (2.32)] The notation for the ansatz, with entries such as L_{1/\bar z,1/\bar z,...} and later 'L ...', is not defined carefully enough to be reproducible. The ellipses and the allowed positions of 1/\bar z should be specified precisely.
  3. [Sec. 3.2.2] The final analytic expressions for I_176 and I_181 are only supplied in ancillary files. Since these are the main claimed results, the paper should display at least the word structure, the first few terms, or a representative piece of each expression so that the claims can be inspected without downloading external files.
  4. [Sec. 2.2] For the four nonplanar two-point master integrals N_1,...,N_4, only the final epsilon expansions are shown. It would improve reproducibility to include the exact topology and the commands used in HyperlogProcedures, so that the reader can independently regenerate these results.
  5. [References] Several references are to online software or AI platforms ([28]-[31], [52]). Please label these consistently as software/online citations, and specify version or access date where relevant.

Circularity Check

1 steps flagged

Bootstrap workflow is mostly self-contained; main residual concern is an explicitly-labeled empirical ansatz rule, not circularity.

specific steps
  1. ansatz smuggled in via citation [Sec. 3.2.1, text near Eq. (3.11)]
    "This function space can be inferred either by the empirical rule in App. B or the new magic identity [50, 51] satisfied by this integral."

    The paper presents two independent routes to the ansatz for I_173: the empirical rule (App. B, explicitly 'only an observation') and the 'new magic identity' from the authors' own prior work [50,51]. The bootstrap result is then fixed within a function space that is justified by citing that same prior work or by an observation explicitly acknowledged as not a principle. This is a mild instance of ansatz support resting on self-citation, but it is not the central claimed derivation: the coefficients are still fixed against independent boundary data and numerically checked with trillo.

full rationale

The paper's central derivation chain is not circular in the strong sense. The leading-singularity decompositions (2.26)–(2.27) are algebraic identities on integrands; the splitting into pieces is exact by construction, with the pieces forming a partition of the original integrand and not defining the result in terms of itself. The boundary data (Sec. 2.2) come from expansion by regions plus four-loop two-point master integrals from Baikov–Chetyrkin and Lee–Smirnov–Smirnov, which are external to this paper; the nonplanar masters are evaluated with HyperlogProcedures and (2.38) provides explicit results. The series-expansion matching (Sec. 2.3) is a linear fitting problem against an independent function space, and the 'mirror constraints' are derived from the same boundary data rather than from the unknown result. The claimed new results for I_176 and I_181 are verified numerically with an independent integrator (trillo, 10^-4 to 10^-5 relative error), which is a genuine external check. The main weakness is a completeness assumption, not a circular definition: App. B states the 'last two entries' rule 'is only an observation rather than a principle', and Sec. 2.1.4 admits the ansatz is 'justified by the success of bootstrap and the verification of numerical values.' This is a possible incompleteness risk (a different function in a larger space could satisfy the same boundary data), and Sec. 2.1.4's own I_22 example shows letters can appear in the last four entries, partially contradicting the rule. But an unproven or even partially contradicted ansatz is a correctness/completeness concern, not circularity: the fitted parameters are not the claimed output, and the output is not fed back into the ansatz. The one self-citation [51] provides the function space for I_173 (a result already obtained in that prior work, not claimed as new here) and is secondary to the paper's main new claims for I_176/I_181. I therefore assign score 2, reflecting the mildly self-cited ansatz support while noting the central derivations are self-contained against independent boundary data and numerical checks.

Axiom & Free-Parameter Ledger

2 free parameters · 5 axioms · 0 invented entities

No new physical entities are introduced. The central calculation depends on an empirical ansatz rule (App. B), the assumption that the relevant integrals are single-valued polylogarithmic, and the sufficiency of truncated boundary data; all are assumptions rather than derived facts. The ansatz coefficients are free parameters fixed by fitting boundary conditions.

free parameters (2)
  • Ansatz coefficients c_i, d_j, e_k = not displayed; fixed by boundary matching
    In Eq. (2.31), the ansatz for I_176^a,b,c is a linear combination with coefficients fixed by matching boundary expansions; for I_173, 304 coefficients are fixed by boundary and mirror constraints. These are fit parameters, not derived from first principles.
  • Split parameter alpha in Eq. (2.27) = taken as 0 (or kept arbitrary)
    The decomposition of I_176 into single-leading-singularity pieces contains an arbitrary constant alpha; the paper sets alpha=0 to obtain I_176. It is a hand-chosen freedom in the construction.
axioms (5)
  • domain assumption Rigid four-point conformal integrals evaluate to single-valued multiple polylogarithms.
    Sec. 2.1.4 states "we expect them to evaluate to svMPLs". This defines the search space; it is not proven for the new integrals and is only checked numerically.
  • ad hoc to paper Empirical rule for ansatz letters: new letters appear only in the last two entries and match leading singularities of reduced graphs.
    App. B explicitly calls it "only an observation rather than a principle". It is load-bearing because it determines which function space is searched.
  • domain assumption Finite conformal integrals have no poles or branch cuts on the real axis, so mirror (inconsistent-limit) constraints are valid.
    Sec. 2.3 uses the "inconsistent" choice of z, zbar to generate extra linear equations based on the claim that the final results have no pole/branch cut on z=zbar.
  • domain assumption Boundary expansions truncated to O(u^0,Y^3) at three loops and O(u^0,Y^4) at four loops, with six boundary limits, suffice to fix all ansatz coefficients.
    Sec. 2.2 and Sec. 2.3; sufficiency is demonstrated only in the worked examples, not proven in general.
  • standard math Gram determinant identities in four dimensions can be used as integral identities.
    Eq. (2.15) and its four-loop generalizations; standard consequence of embedding-space dimension, but used non-trivially to solve I_15^(3) and to reduce integrals.

pith-pipeline@v1.3.0-alltime-deepseek · 26798 in / 13232 out tokens · 125090 ms · 2026-08-02T06:48:49.576189+00:00 · methodology

0 comments
read the original abstract

We set up a bootstrap workflow to study four-point conformal integrals in position space, using leading singularities, single-valued multiple polylogarithmic ans\"atze and boundary data from expansion by regions. These four-point conformal integrals are general in the sense that they are generated by the four-point projections of all possible $f$-graphs, including all non-planar $f$-graph sectors. For three-loop cases, fourteen of the fifteen inequivalent integrand basis can be directly calculated by \texttt{HyperlogProcedures} and the last one is fixed by Gram identity. Then we concentrate on how far the bootstrap workflow can go for four-loop cases, though it works for three-loop cases as well. We show that integrals with several leading singularities can be made tractable by decomposing them into pieces with simpler cut structure. Some four-loop integrals which can not be calculated or very hard to be calculated by other methods for now are obtained in this way. We also provide a package with skill files which is suitable to be read and used by current AI models.

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

54 extracted references · 42 linked inside Pith

  1. [1]

    B. Eden, P. Heslop, G.P. Korchemsky and E. Sokatchev,Hidden symmetry of four-point correlation functions and amplitudes in N=4 SYM,Nucl. Phys. B862(2012) 193 [1108.3557]

    Pith/arXiv arXiv 2012

  2. [2]

    Heslop,The SAGEX Review on Scattering Amplitudes, Chapter 8: Half BPS correlators, J

    P. Heslop,The SAGEX Review on Scattering Amplitudes, Chapter 8: Half BPS correlators, J. Phys. A55(2022) 443009 [2203.13019]

    Pith/arXiv arXiv 2022

  3. [3]

    Schnetz,Graphical functions and single-valued multiple polylogarithms,Commun

    O. Schnetz,Graphical functions and single-valued multiple polylogarithms,Commun. Num. Theor. Phys.08(2014) 589 [1302.6445]

    Pith/arXiv arXiv 2014

  4. [4]

    Borinsky and O

    M. Borinsky and O. Schnetz,Graphical functions in even dimensions,Commun. Num. Theor. Phys.16(2022) 515 [2105.05015]

    Pith/arXiv arXiv 2022

  5. [5]

    Schnetz,Generalized single-valued hyperlogarithms,2111.11246

    O. Schnetz,Generalized single-valued hyperlogarithms,2111.11246

    Pith/arXiv arXiv

  6. [6]

    Schnetz,Graphical functions with spin,JHEP06(2025) 053 [2504.05850]

    O. Schnetz,Graphical functions with spin,JHEP06(2025) 053 [2504.05850]

    Pith/arXiv arXiv 2025

  7. [7]

    Chakraborty, M

    M. Chakraborty, M. Klann, S.-O. Moch, P. Mukherjee, T. Porsche, O. Schnetz et al., Graphical Functions by Examples,2604.25739. – 29 –

    Pith/arXiv arXiv

  8. [8]

    Broadhurst and D

    D.J. Broadhurst and D. Kreimer,Knots and numbers in Phi**4 theory to 7 loops and beyond,Int. J. Mod. Phys. C6(1995) 519 [hep-ph/9504352]

    Pith/arXiv arXiv 1995

  9. [9]

    Schnetz,Quantum periods: A Census of phi**4-transcendentals,Commun

    O. Schnetz,Quantum periods: A Census of phi**4-transcendentals,Commun. Num. Theor. Phys.4(2010) 1 [0801.2856]

    Pith/arXiv arXiv 2010

  10. [10]

    Hyperlogprocedures, version 0.8

    O. Schnetz, “Hyperlogprocedures, version 0.8.” https://www.physik.uni-hamburg.de/th2/ag-moch/personen/schnetz-oliver.html, 2025

    2025

  11. [11]

    Kotikov,Differential equations method: New technique for massive Feynman diagrams calculation,Phys

    A.V. Kotikov,Differential equations method: New technique for massive Feynman diagrams calculation,Phys. Lett. B254(1991) 158

    1991

  12. [12]

    Henn,Multiloop integrals in dimensional regularization made simple,Phys

    J.M. Henn,Multiloop integrals in dimensional regularization made simple,Phys. Rev. Lett. 110(2013) 251601 [1304.1806]

    Pith/arXiv arXiv 2013

  13. [13]

    Caron-Huot and J.M

    S. Caron-Huot and J.M. Henn,Iterative structure of finite loop integrals,JHEP06(2014) 114 [1404.2922]

    Pith/arXiv arXiv 2014

  14. [14]

    Brown,The Massless higher-loop two-point function,Commun

    F. Brown,The Massless higher-loop two-point function,Commun. Math. Phys.287(2009) 925 [0804.1660]

    Pith/arXiv arXiv 2009

  15. [15]

    Panzer,Algorithms for the symbolic integration of hyperlogarithms with applications to Feynman integrals,Comput

    E. Panzer,Algorithms for the symbolic integration of hyperlogarithms with applications to Feynman integrals,Comput. Phys. Commun.188(2015) 148 [1403.3385]

    Pith/arXiv arXiv 2015

  16. [16]

    Kardos, S.-O

    A. Kardos, S.-O. Moch and O. Schnetz,HyperFORM – a FORM package for parametric integration with hyperlogarithms, inLoops and Legs in Quantum Field Theories, 7, 2026 [2607.01163]

    Pith/arXiv arXiv 2026

  17. [17]

    Smirnov,Analytic Tools for Feynman Integrals, vol

    V.A. Smirnov,Analytic Tools for Feynman Integrals, vol. 250 ofSpringer Tracts in Modern Physics, Springer (2012), 10.1007/978-3-642-34886-0

    doi:10.1007/978-3-642-34886-0 2012

  18. [18]

    Chetyrkin and F.V

    K.G. Chetyrkin and F.V. Tkachov,Integration by parts: The algorithm to calculate β-functions in 4 loops,Nucl. Phys. B192(1981) 159

    1981

  19. [19]

    Laporta,High-precision calculation of multiloop Feynman integrals by difference equations, Int

    S. Laporta,High-precision calculation of multiloop Feynman integrals by difference equations, Int. J. Mod. Phys. A15(2000) 5087 [hep-ph/0102033]

    Pith/arXiv arXiv 2000

  20. [20]

    Laporta,Calculation of Feynman integrals by difference equations,Acta Phys

    S. Laporta,Calculation of Feynman integrals by difference equations,Acta Phys. Polon. B 34(2003) 5323 [hep-ph/0311065]

    Pith/arXiv arXiv 2003

  21. [21]

    Drummond, C

    J. Drummond, C. Duhr, B. Eden, P. Heslop, J. Pennington and V.A. Smirnov,Leading singularities and off-shell conformal integrals,JHEP08(2013) 133 [1303.6909]

    Pith/arXiv arXiv 2013

  22. [22]

    Chicherin, T

    D. Chicherin, T. Gehrmann, J.M. Henn, P. Wasser, Y. Zhang and S. Zoia,All Master Integrals for Three-Jet Production at Next-to-Next-to-Leading Order,Phys. Rev. Lett.123 (2019) 041603 [1812.11160]

    Pith/arXiv arXiv 2019

  23. [23]

    Herrmann and J

    E. Herrmann and J. Parra-Martinez,Logarithmic forms and differential equations for Feynman integrals,JHEP02(2020) 099 [1909.04777]

    Pith/arXiv arXiv 2020

  24. [24]

    J. Henn, B. Mistlberger, V.A. Smirnov and P. Wasser,Constructing d-log integrands and computing master integrals for three-loop four-particle scattering,JHEP04(2020) 167 [2002.09492]

    Pith/arXiv arXiv 2020

  25. [25]

    J. Chen, X. Jiang, X. Xu and L.L. Yang,Constructing canonical Feynman integrals with intersection theory,Phys. Lett. B814(2021) 136085 [2008.03045]

    Pith/arXiv arXiv 2021

  26. [26]

    He, Y.-t

    S. He, Y.-t. Huang and C.-K. Kuo,Leading singularities and chambers of Correlahedron, JHEP03(2026) 071 [2505.09808]. – 30 –

    Pith/arXiv arXiv 2026

  27. [27]

    Kuo and Q

    C.-K. Kuo and Q. Yang,Notes on off-shell conformal integrals and correlation functions at five points,Phys. Rev. D113(2026) 105022 [2512.21947]

    Pith/arXiv arXiv 2026

  28. [28]

    Aether: AI for Physics

    Aether Team, “Aether: AI for Physics.”https://aether.aiphys.cn, 2026

    2026

  29. [29]

    DeepSeek

    DeepSeek Team, “DeepSeek.”https://www.deepseek.com, 2026

    2026

  30. [30]

    Gemini Models

    Google DeepMind, “Gemini Models.”https://deepmind.google/models/gemini/, 2026

    2026

  31. [31]

    GLM-5.2 Model

    Zhipu AI, “GLM-5.2 Model.” https://docs.bigmodel.cn/cn/guide/models/text/glm-5.2, 2026

    2026

  32. [32]

    Zhang,Nonplanar integrated correlator in N=4 SYM,Phys

    S.-Q. Zhang,Nonplanar integrated correlator in N=4 SYM,Phys. Rev. D110(2024) 025003 [2404.18900]

    Pith/arXiv arXiv 2024

  33. [33]

    Drummond,Generalised ladders and single-valued polylogarithms,JHEP02(2013) 092 [1207.3824]

    J.M. Drummond,Generalised ladders and single-valued polylogarithms,JHEP02(2013) 092 [1207.3824]

    Pith/arXiv arXiv 2013

  34. [34]

    Dixon, C

    L.J. Dixon, C. Duhr and J. Pennington,Single-valued harmonic polylogarithms and the multi-Regge limit,JHEP10(2012) 074 [1207.0186]

    Pith/arXiv arXiv 2012

  35. [35]

    Chavez and C

    F. Chavez and C. Duhr,Three-mass triangle integrals and single-valued polylogarithms, JHEP11(2012) 114 [1209.2722]

    Pith/arXiv arXiv 2012

  36. [36]

    Drummond, J

    J.M. Drummond, J. Henn, V.A. Smirnov and E. Sokatchev,Magic identities for conformal four-point integrals,JHEP01(2007) 064 [hep-th/0607160]

    Pith/arXiv arXiv 2007

  37. [37]

    Chicherin, A

    D. Chicherin, A. Georgoudis, V. Gon¸ calves and R. Pereira,All five-loop planar four-point functions of half-BPS operators inN= 4SYM,JHEP11(2018) 069 [1809.00551]

    Pith/arXiv arXiv 2018

  38. [38]

    Lee,LiteRed 1.4: a powerful tool for reduction of multiloop integrals,J

    R.N. Lee,LiteRed 1.4: a powerful tool for reduction of multiloop integrals,J. Phys. Conf. Ser.523(2014) 012059 [1310.1145]

    Pith/arXiv arXiv 2014

  39. [39]

    Baikov and K.G

    P.A. Baikov and K.G. Chetyrkin,Four Loop Massless Propagators: An Algebraic Evaluation of All Master Integrals,Nucl. Phys. B837(2010) 186 [1004.1153]

    Pith/arXiv arXiv 2010

  40. [40]

    Lee, A.V

    R.N. Lee, A.V. Smirnov and V.A. Smirnov,Master Integrals for Four-Loop Massless Propagators up to Transcendentality Weight Twelve,Nucl. Phys. B856(2012) 95 [1108.0732]

    Pith/arXiv arXiv 2012

  41. [41]

    Peraro,FiniteFlow: multivariate functional reconstruction using finite fields and dataflow graphs,JHEP07(2019) 031 [1905.08019]

    T. Peraro,FiniteFlow: multivariate functional reconstruction using finite fields and dataflow graphs,JHEP07(2019) 031 [1905.08019]

    Pith/arXiv arXiv 2019

  42. [42]

    Bourjaily, S

    J.L. Bourjaily, S. He, C. Shi and Y. Tang,Four-point correlator of planar supersymmetric Yang-Mills theory at twelve loops,Phys. Rev. D112(2025) 126029 [2503.15593]

    Pith/arXiv arXiv 2025

  43. [43]

    Bourjaily, A

    J.L. Bourjaily, A. DiRe, A. Shaikh, M. Spradlin and A. Volovich,The Soft-Collinear Bootstrap: N=4 Yang-Mills Amplitudes at Six and Seven Loops,JHEP03(2012) 032 [1112.6432]

    Pith/arXiv arXiv 2012

  44. [44]

    Bourjaily, P

    J.L. Bourjaily, P. Heslop and V.-V. Tran,Amplitudes and Correlators to Ten Loops Using Simple, Graphical Bootstraps,JHEP11(2016) 125 [1609.00007]

    Pith/arXiv arXiv 2016

  45. [45]

    S. He, C. Shi, Y. Tang and Y.-Q. Zhang,The cusp limit of correlators and a new graphical bootstrap for correlators/amplitudes to eleven loops,JHEP03(2025) 192 [2410.09859]

    Pith/arXiv arXiv 2025

  46. [46]

    B. Eden, P. Heslop, G.P. Korchemsky and E. Sokatchev,Constructing the correlation function of four stress-tensor multiplets and the four-particle amplitude in N=4 SYM,Nucl. Phys. B862(2012) 450 [1201.5329]. – 31 –

    Pith/arXiv arXiv 2012

  47. [47]

    Fleury and R

    T. Fleury and R. Pereira,Non-planar data ofN= 4 SYM,JHEP03(2020) 003 [1910.09428]

    Pith/arXiv arXiv 2020

  48. [48]

    Bargheer and A

    T. Bargheer and A. Bekov,The Non-Planar Four-Point Integrand and Konishi Dimension in N=4 Super Yang-Mills Theory at Five Loops,2511.21807

    arXiv

  49. [49]

    S. He, C. Shi, Y. Tang and C. Wen,Bootstrapping Giant Graviton Correlators,2605.14281

    Pith/arXiv arXiv

  50. [50]

    Caron-Huot and F

    S. Caron-Huot and F. Coronado,Ten dimensional symmetry ofN= 4 SYM correlators, JHEP03(2022) 151 [2106.03892]

    Pith/arXiv arXiv 2022

  51. [51]

    S. He, X. Jiang, J. Liu and Y.-Q. Zhang,Notes on conformal integrals: Coulomb branch amplitudes, magic identities, and bootstrap,Phys. Rev. D112(2025) 076012 [2502.08871]

    arXiv 2025

  52. [52]

    Tropical integrator that runs inside your browser

    Chen-Yu Wang, Shun-Qing Zhang, “Tropical integrator that runs inside your browser.” https://let-me-trillo-that-for-you.pages.dev, 2026

    2026

  53. [53]

    Schlotterer, Y

    O. Schlotterer, Y. Sohnle and Y.-X. Tao,Elliptic modular graph forms, equivariant iterated integrals and single-valued elliptic polylogarithms,2511.15883

    arXiv

  54. [54]

    Baune, J

    K. Baune, J. Broedel and Y. Moeckli,A construction of single-valued elliptic polylogarithms, 2511.15240. – 32 –

    Pith/arXiv arXiv