Unified web for expansions of amplitudes

Kang Zhou

REVIEW 2 major objections 5 minor 25 references

Unified web for expansions of amplitudes

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper claims that every tree-level amplitude in the unified web—from gravity to special Galileon—can be expanded to double color-ordered bi-adjoint scalar amplitudes, unified in a single double-copy formula derived purely from…

desk verdict A useful organizing result for the amplitudes web, with a plausible but under-verified multi-trace step that a serious referee should ask to be filled in. read the letter →

arxiv 1908.10272 v1 pith:7OIJN53C submitted 2019-08-27 hep-th

Kang Zhou This is my paper · ORCID

classification hep-th

keywords expansiondifferentialoperatorsunifiedwebdoublecopybi-adjointscalarKleiss-KuijfbasisscatteringamplitudesBCJnumerators

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that the web of tree-level amplitude expansions—covering gravity, Einstein-Yang-Mills, Einstein-Maxwell, Born-Infeld, Yang-Mills, Yang-Mills-scalar, $\phi^4$, nonlinear $\sigma$ model, bi-adjoint scalar, Dirac-Born-Infeld, special Galileon, and their variants—is complete and derivable from a single source: the differential-operator relations of Table 1. Applying these operators to known expansions shows that they modify only the coefficients, not the basis. The endpoint is a universal double-copy formula $A = \sum_{\sigma} \sum_{\sigma'} C(\sigma) A_{\mathrm{BAS}}(1,\sigma,n;1,\sigma',n) C(\sigma')$, in which every amplitude in the web is written as a double sum over double color-ordered bi-adjoint scalar amplitudes. If correct, all expansion relations among these theories reduce to bookkeeping of the same numerator coefficients, and the dual web of differential operators and the web of expansions are exactly mapped to each other.

What carries the argument

The machinery is the differential-operator correspondence from Table 1: trace operators $T^{\epsilon}[\bar{\alpha}]$, insertion operators $T^{\epsilon}_{ikj}$, and longitudinal operators $L^{\epsilon}_i, L^{\epsilon}_{ij}$ act on polarization vectors and transmute tree-level gravity amplitudes into amplitudes of the other theories. The paper's move is to let these operators act on the two sides of known expansions: because the operators touch only polarization vectors, they change the coefficients while leaving the KK basis intact. Iterating this procedure generates all expansion arrows in the web, and the double-copy formula (5.8) packages every expansion as a sum over double color-ordered BAS amplitudes with numerator coefficients $C(\sigma)$ obtained by applying the same operators to BCJ numerators.

What would settle it

Compute a five-point NLSM amplitude using the double-copy formula (5.8) with $C_5(\sigma) C_4(\sigma')$ and compare with a direct evaluation of the NLSM amplitude from its integral representation at the same multiplicity; a mismatch in any coefficient of a distinct double color-ordered BAS amplitude would falsify the unified formula. Alternatively, verify the Type-II recursive expansion (3.56) for a four-gluon two-trace EYM amplitude against a Feynman-diagram computation.

Watch

Extended reading notes

Core claim

The central discovery is that the expansions are the dual of the differential-operator map: each operator row of Table 1 corresponds to a row of coefficients in Table 2, and applying operators to expansions yields new expansions without changing the basis. Concretely, the paper derives two recursive expansions for multi-trace EYM amplitudes (Type I with at least one graviton, Type II with only gluons) purely from operator manipulations, then extracts the KK-basis coefficients for GR, EYM, EM, and BI, and applies the remaining operators to obtain expansions for YM, YMS, sYMS, $\phi^4$, NLSM, BAS, exDBI, DBI, and SG. All of these are organized into the unified web of Fig. 1 and unified in the double-copy formula (5.8), where the double color-ordered BAS amplitude supplies the propagator matrix and the coefficients are operator-images of Yang-Mills BCJ numerators.

Load-bearing premise

The web stands on the differential-operator identities of Table 1: that acting with the listed trace, insertion, and longitudinal operators on gravity amplitudes exactly yields the amplitudes of the other theories; if any identity fails off-shell or under some hidden kinematic condition, every expansion built by commuting the operators through known relations inherits the failure.

Editorial extensions

If this is right

Every tree-level amplitude in the web can be expanded algorithmically to double color-ordered bi-adjoint scalar amplitudes, so the BAS theory serves as a universal propagator basis for all the listed theories.
The two new recursive expansions for multi-trace EYM amplitudes reduce any such amplitude to the Kleiss-Kuijf basis of Yang-Mills amplitudes at tree level.
Because all expansions share the same KK basis, coefficients for different theories are related by the same differential operators that relate the amplitudes themselves (Eq. (4.4)).
The generalized KK relations of Section 2.2 hold for YMS, sYMS, $\phi^4$, NLSM, and BAS amplitudes, not just Yang-Mills.
The double-copy formula gives a concrete numerator construction for DBI, SG, and other effective-field-theory amplitudes: choose the Table-2 coefficients and sum over BAS amplitudes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A practical consequence not spelled out by the paper: for any target theory in the web, amplitudes at higher multiplicity can be computed by summing BAS amplitudes (pure propagators) weighted by known BCJ-type numerators, which could outperform direct Feynman-diagram evaluation for DBI or SG.
The duality between Table 1 and Table 2 suggests that any new differential-operator identity that maps GR to a different theory would automatically produce a new expansion to BAS amplitudes, extending the web beyond its current nodes.
The paper's closing question—what theories correspond to mixed coefficient pairs such as $C^{\epsilon}_2(\sigma)C^{\tilde{\epsilon}}_{3b}(\sigma')$—is directly testable: evaluating the double sum (5.8) for such a pair would define a new amplitude and its integral representation.
If the operator identities of Table 1 survive at loop level, the same web structure and the same coefficient relations might carry over to loop integrands, though the paper is strictly tree-level.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Kang Zhou has put together a genuinely useful organizing result: a single web of expansions connecting a dozen theories, with a universal double copy formula A = sum C(σ) A_BAS C(σ') at the end. The main new content is the recursive expansion of multi-trace EYM amplitudes to a YM basis, derived by pushing trace operators through the single-trace expansion. I think the core idea is right, and the paper is honest about where it stands on previous work: the operator web comes from Cheung-Shen-Wen [8], the single-trace expansion from Feng-Li-Zhou [21], and multi-trace expansions from CHY already appeared in Du-Feng-Teng [20]. What is added is the differential-operator derivation and the complete web, which is a clean conceptual package rather than a new calculational tool.

Where I would push back: the generalization from double trace to arbitrary multi-trace, Eqs. (3.42) and (3.56), is stated as "the calculation is exactly the same" and then given in compressed notation. That is not a full derivation. The notation K_{b,a} treated as a single element in shuffles and T^{mu nu} = k_a^mu k_b^nu is clever, but I share the stress-test's concern that independence of the order in which traces are processed is not demonstrated. The double-trace case is worked out in detail; the step from there to r traces involves nested sums over subsets, endpoints, and ordered sets, and a sign or double-counting error there would indeed change the coefficients. The paper gives no numerical checks of (3.40), (3.42), (3.55), or (3.56), so the completeness claim of the web rests on a plausible but unverified combinatorial identity. This is a normal thing to ask an author to supply in revision; it is not a reason to reject.

Also minor: section 4.1 explicitly acknowledges coefficients are not unique due to BCJ relations, which is fine, but it means the relations in (4.4) hold for a particular set of coefficients and the discussion is a bit loose. The algorithms in section 4.2 are cited from the literature, not new, and are fine.

Bottom line: this deserves a serious referee. The right outcome is probably a major-revision request to flesh out the multi-trace derivation and add at least one numerical or symbolic check of (3.42)/(3.56). I would cite it as a useful map of the expansion web.

Referee Report

2 major / 5 minor

Summary. This paper constructs a 'unified web' of tree-level amplitude expansions. Starting from the differential-operator transmutations collected in Table 1 (from Cheung-Shen-Wen), the author re-derives the recursive expansions of single-trace EYM and GR amplitudes to color-ordered YM amplitudes, then derives new recursive expansions for multi-trace EYM amplitudes (Type-I with a graviton, Type-II without), extracts KK-basis coefficients for GR, EYM, EM, and BI, and by acting with the complementary polarization operators obtains expansions of YM, YMS, sYMS, φ4, NLSM, BAS, exDBI, DBI, and SG. All expansions are assembled into the universal double-copy formula A = Σ_σ Σ_σ' C(σ) A_BAS(1,σ,n;1,σ',n) C(σ') (Eq. 5.8). The paper claims that the whole web follows from Table 1 alone, with no additional assumptions and no fitted parameters.

Significance. If the new multi-trace EYM expansions are correct, the paper achieves a genuine unification: every expansion in the web is obtained from one double-copy formula, and there is an explicit duality between differential operators and expansion coefficients. The manuscript is self-consistent, contains no free parameters, gives a transparent review of the coefficient algorithms in Section 4, and explicitly acknowledges the BCJ-related non-uniqueness of coefficients in Section 4.1. These are real strengths. The main caveat is that the load-bearing new input—the multi-trace EYM formulas (3.42) and (3.56)—is stated rather than fully derived and is not checked numerically or against the known CHY-based results of [20]. Until that support is supplied, the 'complete unified web' claim remains conditional.

major comments (2)

[§3.1, Eq. (3.42)] The transition from the detailed two-trace calculation to the arbitrary multi-trace Type-I expansion is asserted ('The calculation is exactly the same') rather than proved. Equation (3.42) involves a subset TrTrs of traces, paired endpoints a_i,b_i, and the ordered set K̄(TrTrs,a,b), with each K^{t_i}_{b_i,a_i} treated as a single element in a shuffle. The paper does not demonstrate that applying the commuting trace operators in different orders yields the same sum, nor that configurations with several K sets adjacent in the chain are counted exactly once. Since the coefficients C~ϵ_2(σ) obtained from this expansion feed Eqs. (4.2), (4.4), and ultimately the universal double-copy formula (5.8), an ordering or sign error here would invalidate the central claim. Please supply an induction proof or, at minimum, explicit low-multiplicity checks (e.g., 6-point double- and triple-trace examples) compared with the CHY-based results of [20].
[§3.2, Eqs. (3.55)–(3.58)] The no-graviton multi-trace expansion is the least supported formula in the paper. Equation (3.55) is derived only in compressed words, and the jump to the general formula (3.56) is stated as 'can be obtained directly'. The coefficient Ĉ^{d2}_{a,b} in (3.58) assigns T^{μν} = k_a^μ k_b^ν to each K set and treats K sets as single elements in the shuffle, but the paper does not show the sign and momentum-flow assignments when several K sets appear together with K^{222}_{d2,c2} in one chain. The exclusion of cases where β1 or βr lies outside {α,c2} is also asserted rather than proved. This formula is not compared with the corresponding result of [20], despite the acknowledged difference in classification. Please provide at least the triple-trace case in detail or a numerical cross-check.

minor comments (5)

[§2.2, Eq. (2.17)] In the first generalized KK relation, the left-hand YMS amplitude is written with a graviton set HHH, but YMS amplitudes in Table 1 contain gluons GGG in the second ordering; as printed, the two sides of the equality are inconsistent.
[§2.3] The statement that the basis and recursive expansions 'can be obtained only through knowledge of differential operators' is stronger than what is demonstrated: the paper shows that the KK basis emerges naturally from the trace operators, but it does not prove uniqueness of that basis choice among all possible complete bases.
[Figure 1] The line types and arrow directions in the web are hard to distinguish in a grayscale print; a table or a legend explicitly listing which coefficient each line type denotes would improve readability.
[Throughout] There are numerous language errors and typos, e.g., 'thoes' in the introduction, 'dived' in Eq. (2.3), 'the subscribe' in Section 2.1, and 'visitable' in Section 6; a careful proofreading pass is needed.
[§4.2] The notation for ordered splittings uses the same symbol Z for combinatory momenta as for the set of external particles in Section 2, and ρ is used both for pair partitions and for elements of an ordered sequence; using distinct symbols would avoid confusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all expansions follow from external differential-operator identities and prior parameter-free derivations; the double-copy formula is a derived repackaging, not an input.

full rationale

The derivation chain is self-contained modulo the external differential-operator identities of Table 1 [8]. The single-trace EYM and GR recursive expansions (2.20)/(2.26) are cited to [21], which contains the present author, but [21] is a parameter-free derivation from those identities plus gauge invariance and does not assume the multi-trace expansions or the double-copy formula being proved here; it is therefore independent support rather than a circular premise. The multi-trace EYM expansions (3.40)/(3.42)/(3.56) are obtained by applying fixed trace operators to the single-trace expansion, not by assuming the target coefficients. The coefficient relations (4.4) follow by linearity because the ~epsilon-operators act only on the ~epsilon-dependent coefficients and annihilate the epsilon-polarized YM basis, and the expansions in Sec. 5 are obtained by applying epsilon-operators to both sides, leaving coefficients unchanged. The universal double-copy formula (5.8) is a direct consequence of substituting the BAS expansion of color-ordered YM amplitudes into the GR expansion; it is a derived repackaging theorem, not a fitted prediction and not an input assumed from the authors' prior work. No parameter is fit, no target amplitude is used as an input, and no uniqueness theorem from the authors' prior papers is invoked to forbid alternatives. The unverified combinatorial details of (3.42)/(3.56) would be a correctness risk, not a circularity, since they are derived rather than assumed.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No numbers are fitted to data or chosen by hand; the paper is a derivation of analytic relations with no free parameters. No new particles, forces, or fields are introduced; the web is a conceptual organization of existing amplitudes. The axioms correspond to the external differential-operator framework and the known recursive expansions that the paper takes as input.

assumptions (3)

domain assumption Differential operator relations in Table 1 transmute tree-level GR amplitudes to amplitudes of EYM, EM, BI, YM, YMS, φ4, NLSM, BAS, DBI, and SG.
Taken from the external work [8]; the paper states it uses only these relations as input to construct all expansions.
domain assumption The recursive expansions for single-trace EYM and GR amplitudes (Eqs. 2.20 and 2.26) are valid.
Derived in the self-cited work [21] by solving differential equations and using gauge invariance; the current paper applies trace operators to these expansions to obtain multi-trace results.
standard math Generalized Kleiss-Kuijf relations hold for the color-ordered amplitudes of the theories considered.
Used in Section 2.2 to define the KK basis; the paper derives them from the algebraic properties of the trace operator, so this is background linear algebra of the operator framework.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Unified web for expansions of amplitudes." pith.science (2026). https://pith.science/paper/7OIJN53C

@misc{pith2026190810272,
  author       = {Pith},
  title        = {Pith review of: Unified web for expansions of amplitudes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7OIJN53C}},
  note         = {Machine review of arXiv:1908.10272}
}

abstract

In this paper, we demonstrate that using differential operators one can construct the complete unified web for expansions of amplitudes for a wide range of theories. We first re-derive the expansion of multi-trace Einstein-Yang-Mills amplitudes to Kleiss-Kuijf basis of color-ordered Yang-Mills amplitudes, by applying proper differential operators which modify the coefficients in the recursive expansion of single-trace Einstein-Yang-Mills amplitudes. Next, through differential operators which act on amplitudes only, we obtain expansions of amplitudes of Yang-Mills theory, Yang-Mills-scalar theory, $\phi^4$ theory, non-linear sigma model, bi-adjoint scalar theory, Born-Infeld theory, Dirac-Born-Infeld theory and special Galileon theory. Then, together with other results in literatures, the complete unified web is achieved. This web for expansions is the dual version of the unified web for differential operators. Thus, connections among amplitudes of a variety of theories, which are reflected by Cachazo-He-Yuan integrands and differential operators previously, can also be represented by expansions. We also find that amplitudes of all theories in the web can be expanded to double color-ordered bi-adjoint scalar amplitudes in the universal double copy formula.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 3 canonical work pages

[8]

Cheung, C

C. Cheung, C. H. Shen and C. Wen, JHEP 1802, 095 (2018) [arXiv:1705.03025 [hep-th]]

arXiv 2018
[21]

B. Feng, X. Li and K. Zhou, arXiv:1904.05997 [hep-th]

arXiv 1904
[20]

Y. J. Du, B. Feng and F. Teng, JHEP 1712, 038 (2017) [arXiv:1708.04514 [hep-th]]. – 30 –

arXiv 2017
[1]

Kawai, D

H. Kawai, D. C. Lewellen and S. H. Tye, Nucl. Phys. B 269, 1 (1986)

work page 1986
[2]

Z. Bern, J. J. M. Carrasco and H. Johansson, Phys. Rev. D 78, 085011 (2008) [arXiv:0805.3993 [hep-ph]]

arXiv 2008
[3]

Cachazo, S

F. Cachazo, S. He, and E. Y. Yuan, Phys. Rev. D90 (2014) no. 6, 065001, arXiv:1306.6575 [hep-th]

arXiv 2014
[4]

Cachazo, S

F. Cachazo, S. He, and E. Y. Yuan, Phys. Rev. Lett. 113 (2014) no. 17, 171601, arXiv:1307.2199 [hep-th]

arXiv 2014
[5]

Cachazo, S

F. Cachazo, S. He, and E. Y. Yuan, JHEP 1407 (2014) 033, arXiv:1309.0885 [hep-th]

arXiv 2014

Show all 25 references

[6]

Cachazo, S

F. Cachazo, S. He and E. Y. Yuan, JHEP 1501, 121 (2015) [arXiv:1409.8256 [hep-th]]

2015 arXiv
[7]

Cachazo, S

F. Cachazo, S. He and E. Y. Yuan, JHEP 1507, 149 (2015) [arXiv:1412.3479 [hep-th]]

2015 arXiv
[9]

Zhou and B

K. Zhou and B. Feng, JHEP 1809, 160 (2018) [arXiv:1808.06835 [hep-th]]

2018 arXiv
[10]

Bollmann and L

M. Bollmann and L. Ferro, JHEP 1901, 180 (2019) [arXiv:1808.07451 [hep-th]]

2019 arXiv
[11]

Stieberger and T

S. Stieberger and T. R. Taylor, Nucl. Phys. B 913, 151 (2016) [arXiv:1606.09616 [hep-th]]

2016 arXiv
[12]

Schlotterer, JHEP 1611, 074 (2016) [arXiv:1608.00130 [hep-th]]

O. Schlotterer, JHEP 1611, 074 (2016) [arXiv:1608.00130 [hep-th]]

2016 arXiv
[13]

Chiodaroli, M

M. Chiodaroli, M. Gunaydin, H. Johansson and R. Roiban, JHEP 1707, 002 (2017) doi:10.1007/JHEP07(2017)002 [arXiv:1703.00421 [hep-th]]

2017 arXiv
[14]

Del Duca, L

V. Del Duca, L. J. Dixon and F. Maltoni, Nucl. Phys. B 571, 51 (2000) doi:10.1016/S0550-3213(99)00809-3 [hep-ph/9910563]

2000 arXiv
[15]

Nandan, J

D. Nandan, J. Plefka, O. Schlotterer and C. Wen, JHEP 1610, 070 (2016) [arXiv:1607.05701 [hep-th]]

2016 arXiv
[16]

de la Cruz, A

L. de la Cruz, A. Kniss and S. Weinzierl, Phys. Lett. B 767, 86 (2017) [arXiv:1607.06036 [hep-th]]

2017 arXiv
[17]

C. H. Fu, Y. J. Du, R. Huang and B. Feng, JHEP 1709, 021 (2017) [arXiv:1702.08158 [hep-th]]

2017 arXiv
[18]

Teng and B

F. Teng and B. Feng, JHEP 1705, 075 (2017) [arXiv:1703.01269 [hep-th]]

2017 arXiv
[19]

Y. J. Du and F. Teng, JHEP 1704, 033 (2017) [arXiv:1703.05717 [hep-th]]

2017 arXiv
[22]

S. Q. Hu and K. Zhou, arXiv:1907.07857 [hep-th]

1907 arXiv
[23]

Kleiss and H

R. Kleiss and H. Kuijf, Nucl. Phys. B 312, 616 (1989)

1989
[24]

Baadsgaard, N

C. Baadsgaard, N. E. J. Bjerrum-Bohr, J. L. Bourjaily and P. H. Damgaard, JHEP 1509, 129 (2015) doi:10.1007/JHEP09(2015)129 [arXiv:1506.06137 [hep-th]]

2015 arXiv
[25]

Baadsgaard, N

C. Baadsgaard, N. E. J. Bjerrum-Bohr, J. L. Bourjaily and P. H. Damgaard, JHEP 1509, 136 (2015) doi:10.1007/JHEP09(2015)136 [arXiv:1507.00997 [hep-th]]. – 31 –

2015 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [8]

Cheung, C

C. Cheung, C. H. Shen and C. Wen, JHEP 1802, 095 (2018) [arXiv:1705.03025 [hep-th]]

arXiv 2018

[2] [21]

B. Feng, X. Li and K. Zhou, arXiv:1904.05997 [hep-th]

arXiv 1904

[3] [20]

Y. J. Du, B. Feng and F. Teng, JHEP 1712, 038 (2017) [arXiv:1708.04514 [hep-th]]. – 30 –

arXiv 2017

[4] [1]

Kawai, D

H. Kawai, D. C. Lewellen and S. H. Tye, Nucl. Phys. B 269, 1 (1986)

work page 1986

[5] [2]

Z. Bern, J. J. M. Carrasco and H. Johansson, Phys. Rev. D 78, 085011 (2008) [arXiv:0805.3993 [hep-ph]]

arXiv 2008

[6] [3]

Cachazo, S

F. Cachazo, S. He, and E. Y. Yuan, Phys. Rev. D90 (2014) no. 6, 065001, arXiv:1306.6575 [hep-th]

arXiv 2014

[7] [4]

Cachazo, S

F. Cachazo, S. He, and E. Y. Yuan, Phys. Rev. Lett. 113 (2014) no. 17, 171601, arXiv:1307.2199 [hep-th]

arXiv 2014

[8] [5]

Cachazo, S

F. Cachazo, S. He, and E. Y. Yuan, JHEP 1407 (2014) 033, arXiv:1309.0885 [hep-th]

arXiv 2014