betaFeedback

WorksPhysics and Astronomy

Quadratic Sum Rules for Scalar Amplitudes

Abstract

Kleiss–Kuijf and Bern–Carrasco–Johansson relations, and their string monodromy generalizations, have long provided some of the most powerful and useful constraints on partial amplitudes. Motivated by their success, we uncover a new class of quadratic-sum rules that connect scalar amplitudes at different multiplicities. Combining recursion, kinematic shifts, and symbolic regression, we establish them for the nonlinear sigma model and its string-inspired extensions. We then extend these relations to diagonal elements of the inverse Kawai–Lewellen–Tye matrix and subsequently to planar scalar theories with nonderivative interactions and general propagators.

Introduction

Recent developments have revealed close connections among colored scalar theories, including the non-linear sigma model (NLSM) and bi-adjoint scalars (BAS). Many essential insights have emerged from combining them into larger unified frameworks. Early works on the NLSM derived recursion relations for its amplitudes using the vanishing soft limit [1, 2, 3, 4]. Subsequent work revealed that the soft limits contain “mixed amplitudes”, where some of the pions are replaced with bi-adjoint scalars [5]. This naturally leads to an extended theory containing both pions and BAS, which provides a natural completion of pion amplitudes to odd multiplicities and admits a set of very efficient recursion relations [6].

More recent research shows that these scalar amplitudes can be deformed into one another through kinematic shifts. The most well-known example is the δ\delta-shift [7, 8], which takes the diagonal BAS amplitudes, known as Tr⁡ϕ3\operatorname{Tr}\phi^3 amplitudes, and transmutes them to NLSM or mixed amplitudes at even and odd points, respectively. While a powerful tool, this process is not invertible, and not all mixed amplitudes can be obtained through this procedure. These shortcomings can be overcome by considering string-inspired α′\alpha'-completions of both theories. Historically the most studied α′\alpha'-completion of NLSM is semi-abelianized ZZ-theory [9]. However, more recent work has focused on abelianizing the inverse Kawai–Lewellen–Tye (KLT) kernel [10, 11]. The latter construction admits a different kinematical deformation, the α′\alpha'-shift, which provides an invertible transformation between inverse KLT and stringy pions, and hence demonstrates complete equivalence of both theories. These unifications are illustrated in Figure 1.

Figure 1. Unity of colored scalar theories.

In this work, we uncover quadratic-sum rules among the amplitudes within this extended framework. Consistency conditions between Berends–Giele-type recursions [12, 13] and the δ\delta-shift reveal an initial set of sum rules involving pion and mixed amplitudes. This prompts us to use symbolic regression to discover complementary pion sum rules, which we further extend to a set of sum rules valid for their α′\alpha'-completions. Using the unity of scalar theories, we derive an equivalent statement for inverse

KLT elements. A functional proof of these statements highlights applicability to a broad range of scalar theories without derivative interactions, which can be summarized by the formula

∑k=2n−1[Δ(X1k)−Δ(Xkn)]A(1…k)A(k…n)=0.(1)\sum_{k=2}^{n-1}\left[\Delta(X_{1k})-\Delta(X_{kn})\right]A(1\ldots k)A(k\ldots n)=0. \tag*{(1)}

Here Δ(p2)\Delta(p^2) is a propagator in momentum space, and AA denotes a color-ordered amplitude. We provide several proofs of these sum rules. We also discuss generalizations of these sum rules to theories with spin, derivative interactions, and strings.

Let us compare our sum rules with other relations known in the literature. In the context of scattering amplitudes, the most prominent relations are the linear Kleiss–Kuijf [14] and Bern–Carrasco–Johansson [15] relations. Their power is illustrated, for example, in NLSM, where they reduce the tree-level amplitude basis [16] and help constrain both the leading-order amplitudes [17] and their higher-derivative corrections [18, 19]. Beyond these linear relations, quadratic relations have also been identified, although in more specialized settings. Examples include relations among gauge-theory amplitudes [20] at fixed multiplicities and, more recently, relations for NLSM currents [21].

Review of Colored Scalars

Notation and Conventions. We will be interested in tree-level color-ordered scattering amplitudes in scalar theories. The natural kinematic variables to describe these amplitudes are the planar Mandelstam variables

Xij=(pi+pi+1+⋯+pj−1)2.(2)X_{ij} = (p_i + p_{i+1} + \cdots+ p_{j-1})^2. \tag*{(2)}

These XijX_{ij}-variables label the poles of partial amplitudes and form a basis for kinematic space in general dimensions. Since the ordering of momenta is fixed, we can interpret the amplitudes abstractly as functions of particle labels. Concretely, for some ordering σ\sigma on nn particles, we denote

m(σ)≡m(1,2,…,n)∣Xij→Xσ(i)σ(j).(3)m(\sigma) \equiv m(1,2,\ldots,n)\big|_{X_{ij}\to X_{\sigma(i)\sigma(j)}}. \tag*{(3)}

The advantage of viewing amplitudes as functions of labels rather than of momenta is that the XijX_{ij} may be treated as independent formal variables. This allows us to interpret the same expressions as evaluated on higher-point kinematics, and we can use the same notation for off-shell objects such as loop integrands.

We will make use of this notational freedom in one important way. Some of the formulas below will contain boundary terms in which two-point functions appear together with their propagator. Setting m(i,i+1)=Xi i+1m(i,i+1)=X_{i\,i+1}, both factors will vanish in the on-shell limit, but their ratio approaches unity1^{1}. This will help streamline various formulas into a single sum, which would otherwise require isolating the two-point cases into separate terms.

The main objects we will encounter are NLSM, mixed, and Tr⁡ϕ3\operatorname{Tr}\phi^3 amplitudes. For further clarity, we shall denote pure pion amplitudes by AA, mixed amplitudes by MM, and Tr⁡ϕ3\operatorname{Tr}\phi^3 amplitudes by mm. Throughout this text, we exclusively consider mixed amplitudes with three adjacent scalars placed on labels 1,n−1,n1,n-1,n, or, explicitly, M(ϕπ…πϕϕ)M(\phi\pi\ldots\pi\phi\phi).

α′\alpha'-Completions. We will make use of certain stringy α′\alpha'-extensions of the above field theory amplitudes, which contain these amplitudes in the α′→0\alpha'\to0 limit. The specific class of α′\alpha'-extensions is motivated by the inverse KLT kernel, first studied by Mizera in [22, 23], where it was shown that its elements are simple functions of Mandelstam variables. For the purposes of this text, we only consider the diagonal elements, which we denote mα′(1…n)≡mα′(1∣1)m^{\alpha'}(1\ldots n)\equiv m^{\alpha'}(1|1). Explicit examples of these functions are

mα′(123)=1,mα′(1234)=1t13+1t24,(4)m^{\alpha'}(123)=1,\qquad m^{\alpha'}(1234)=\frac{1}{t_{13}}+\frac{1}{t_{24}}, \tag*{(4)}
mα′(12345)=1+(1t13t35+cyc.),(5)m^{\alpha'}(12345)=1+\left(\frac{1}{t_{13}t_{35}}+\text{cyc.}\right), \tag*{(5)}

where we use the notation tij≡tan⁡(α′πXij)t_{ij}\equiv\tan(\alpha'\pi X_{ij}), and +cyc+\text{cyc} denotes the sum over all five cyclic permutations.

It is possible to obtain stringy NLSM amplitudes from the inverse KLT elements via the α′\alpha'-shift [11]

α′→α′/2,Xeven→Xeven±1/α′.(6)\alpha'\to\alpha'/2,\qquad X_{\mathrm{even}}\to X_{\mathrm{even}}\pm1/\alpha'. \tag*{(6)}

Here, XevenX_{\mathrm{even}} specifies XijX_{ij} with ∣j−i∣|j-i| even. Examples of these amplitudes are

Aα′(1234)=−τ13−τ24,(7)A^{\alpha'}(1234)=-\tau_{13}-\tau_{24}, \tag*{(7)}
Mα′(12345)=1−τ13+τ24τ14−τ24+τ35τ25+τ13τ35,(8)M^{\alpha'}(12345)=1-\frac{\tau_{13}+\tau_{24}}{\tau_{14}}-\frac{\tau_{24}+\tau_{35}}{\tau_{25}}+\tau_{13}\tau_{35}, \tag*{(8)}

where we have defined τij≡tan⁡(α′πXij2)\tau_{ij}\equiv\tan\left(\alpha'\pi\frac{X_{ij}}{2}\right). We stress that the α′\alpha'-shift is reversible, which will be essential later in this text.

To close this section, let us mention that for field theory NLSM amplitudes it is sufficient to start with Tr⁡ϕ3\operatorname{Tr}\phi^3 amplitudes and employ δ\delta-shift defined as [7, 8]

Xo,o→Xo,o−δ,Xe,e→Xe,e+δ,(9)X_{o,o}\to X_{o,o}-\delta,\qquad X_{e,e}\to X_{e,e}+\delta, \tag*{(9)}

where Xo,o,Xe,eX_{o,o},X_{e,e} are XX-variables with any two odd and any two even labels, respectively. The NLSM amplitudes are then obtained as limits as δ\delta goes to infinity:

lim⁡δ→∞δ2n−2m2n(δ)=A2n,(10)\lim_{\delta\to\infty}\delta^{2n-2}m_{2n}(\delta)=A_{2n}, \tag*{(10)}
lim⁡δ→∞δ2n−2m2n+1(δ)=M2n+1.(11)\lim_{\delta\to\infty}\delta^{2n-2}m_{2n+1}(\delta)=M_{2n+1}. \tag*{(11)}

In other words, by taking δ\delta to infinity, we recover pion amplitudes at even-point and mixed amplitudes at odd-point kinematics. Unlike its stringy cousin, the δ\delta-shift is generally not invertible.

Mixed Amplitudes Sum Rules

We begin our exploration by considering the Berends–Giele recursion for even-point Tr⁡ϕ3\operatorname{Tr}\phi^3 amplitudes [6], to which we apply the δ\delta-shift (9). Expanding around large δ\delta yields

m2n(δ)=δ3−2nS1+δ2−2nS2+O(δ1−2n).(12)m_{2n}(\delta)=\delta^{3-2n}S_1+\delta^{2-2n}S_2+O(\delta^{1-2n}). \tag*{(12)}

Comparing against known asymptotic behavior (10) we see that S2S_2 is the expected pion amplitude, whereas S1S_1 has to vanish. Writing S1S_1 explicitly yields the first sum rule

∑k evenA(1…k)M(k…n)X1,k−∑k oddM(1…k)A(k…n)Xk,n=0.(13)\sum_{k\ \mathrm{even}}\frac{A(1\ldots k)M(k\ldots n)}{X_{1,k}}-\sum_{k\ \mathrm{odd}}\frac{M(1\ldots k)A(k\ldots n)}{X_{k,n}}=0. \tag*{(13)}

A simple example for n=6n=6 reads

M(12345)−M(23456)=A(1234)X14−A(3456)X36.(14)M(12345)-M(23456)=\frac{A(1234)}{X_{14}}-\frac{A(3456)}{X_{36}}. \tag*{(14)}

The natural next step is to consider m2n+1(δ)m_{2n+1}(\delta), however, its δ\delta expansion has no super-leading terms and hence does not yield an additional sum rule. Consequently, the presented sum rule only covers the case of even-point kinematics where the highest multiplicity amplitudes present are mixed amplitudes.

Symbolic Regression and Pion Sum Rules

In the previous section we have described an analytical derivation of a sum rule for mixed amplitudes. Although the same line of reasoning fails to provide a pion sum rule, we may nevertheless look for one. While an exhaustive search over all rational functions is intractable, recent developments in machine learning provide a potential way forward: symbolic regression. Given the previous expression, we are motivated to look for A(1…n−1)−A(2…n)A(1\ldots n-1)-A(2\ldots n) as a simple function of lower point amplitudes and XX-variables. Here, we proceed by employing symbolic regression and its implementation in PySR [24]. Symbolic regression has seen use in the context of physics as a tool for rediscovering physical laws, a popular example being AI Feynman [25]. For a recent scattering amplitudes application see [26], where symbolic regression is used to rediscover the double copy relation of gravity and Yang-Mills.

In our case, we are looking for

A(1…n−1)−A(2…n)=F(A,M,X,×,+,−,÷).(15)A(1\ldots n-1)-A(2\ldots n)=F(A,M,X,\times,+,-,\div). \tag*{(15)}

That is, we are looking to express the left-hand side as a rational function in lower-point pion and mixed amplitudes and XX-variables, with the mass dimension fixed to match pion amplitudes.

Indeed, we find the desired sum rule for pion amplitudes

∑k odd(Xk,n−X1,k)M(1…k)M(k…n)+∑k even(Xk,n−X1,k)A(1…k)A(k…n)X1,kXk,n=0.(16)\sum_{k\ \mathrm{odd}}(X_{k,n}-X_{1,k})M(1\ldots k)M(k\ldots n)+\sum_{k\ \mathrm{even}}(X_{k,n}-X_{1,k})\frac{A(1\ldots k)A(k\ldots n)}{X_{1,k}X_{k,n}}=0. \tag*{(16)}

Symbolic regression, with some manual help, discovered these expressions up to n=9n=9. An explicit example for n=7n=7 reads

A(123456)−A(234567)=(X57−X15)M(12345)+(X37−X13)M(34567)+(X47−X14)A(1234)A(4567)X14X47.(17)A(123456)-A(234567)=(X_{57}-X_{15})M(12345)+(X_{37}-X_{13})M(34567)+(X_{47}-X_{14})\frac{A(1234)A(4567)}{X_{14}X_{47}}. \tag*{(17)}

KLT Sum Rules and α′\alpha'-Completion

So far we have only considered field-theory amplitudes. However, as explained in the introduction, α′\alpha'-completion has proven to be an incredibly useful tool and so we may be inclined towards extending the previous results to stringy amplitudes. Our approach is very similar to that of the previous section: we employ symbolic regression with only two changes. First, the functions under consideration are stringy NLSM and mixed amplitudes, defined by α′\alpha'-shift (6). Secondly, we replace XX-variables with both tijt_{ij} and τij\tau_{ij}. Symbolic regression now provides two sets of sum rules

∑k evenAα′(1…k)Mα′(k…n)(1τ1k+τkn)−∑k oddMα′(1…k)Aα′(k…n)(τ1k+1τkn)=0(18)\sum_{k\ \mathrm{even}}A^{\alpha'}(1\ldots k)M^{\alpha'}(k\ldots n)\left(\frac{1}{\tau_{1k}}+\tau_{kn}\right)-\sum_{k\ \mathrm{odd}}M^{\alpha'}(1\ldots k)A^{\alpha'}(k\ldots n)\left(\tau_{1k}+\frac{1}{\tau_{kn}}\right)=0 \tag*{(18)}

for nn even and

∑k odd(τk,n−τ1,k)Mα′(1…k)Mα′(k…n)+∑k even(1τ1,k−1τk,n)Aα′(1…k)Aα′(k…n)=0(19)\sum_{k\ \mathrm{odd}}(\tau_{k,n}-\tau_{1,k})M^{\alpha'}(1\ldots k)M^{\alpha'}(k\ldots n)+\sum_{k\ \mathrm{even}}\left(\frac{1}{\tau_{1,k}}-\frac{1}{\tau_{k,n}}\right)A^{\alpha'}(1\ldots k)A^{\alpha'}(k\ldots n)=0 \tag*{(19)}

for nn odd. Again, let us give the basic examples

Aα′(1234)−Aα′(2345)=(τ35−τ13)Mα′(123)Mα′(345),A^{\alpha'}(1234)-A^{\alpha'}(2345)=(\tau_{35}-\tau_{13})M^{\alpha'}(123)M^{\alpha'}(345),
Mα′(12345)−Mα′(23456)=Aα′(1234)τ14−Aα′(3456)τ36+τ13τ35−τ24τ46.(20)M^{\alpha'}(12345)-M^{\alpha'}(23456)=\frac{A^{\alpha'}(1234)}{\tau_{14}}-\frac{A^{\alpha'}(3456)}{\tau_{36}}+\tau_{13}\tau_{35}-\tau_{24}\tau_{46}. \tag*{(20)}

for the pion and the mixed amplitudes, respectively. Now we recall the α′\alpha'-shift (6) and immediately observe that under its reversal, the sum rules (18), (19) are mapped to

∑k=2n−1(1t1k−1tkn)mα′(1…k)mα′(k…n)=0.(21)\sum_{k=2}^{n-1}\left(\frac{1}{t_{1k}}-\frac{1}{t_{kn}}\right)m^{\alpha'}(1\ldots k)m^{\alpha'}(k\ldots n)=0. \tag*{(21)}

Thus, we obtain a sum rule for the inverse KLT elements. A simple example for n=6n=6 reads

mα′(12345)−mα′(23456)=(1t13−1t36)mα′(3456)+(1t14−1t46)mα′(1234).(22)m^{\alpha'}(12345)-m^{\alpha'}(23456)=\left(\frac{1}{t_{13}}-\frac{1}{t_{36}}\right)m^{\alpha'}(3456)+\left(\frac{1}{t_{14}}-\frac{1}{t_{46}}\right)m^{\alpha'}(1234). \tag*{(22)}

By virtue of the reversibility of the α′\alpha'-shift, all these α′\alpha'-complete sum rules are equivalent to one another. Under consideration of the α′→0\alpha'\to0 limit, these stringy relations then imply the field theory sum rules for pions, mixed amplitudes, and Tr ϕ3\mathrm{Tr}\,\phi^3. What remains is to provide a proof of the KLT sum rule for pions, and the rest follows. The relevant proof is provided in a dedicated Appendix.

Extending the sum rules

The presented sum rules can be extended to a much broader class of theories. Let us consider a theory given by the Lagrangian

L=−12⟨ϕΔ−1ϕ⟩+V(ϕ),(23)\mathcal{L}=-\frac{1}{2}\langle\phi\Delta^{-1}\phi\rangle+V(\phi), \tag*{(23)}

where ϕ=ϕaTa\phi= \phi_a T^a and ⟨A⟩≡Tr⁡A\langle A\rangle\equiv\operatorname{Tr} A. The only essential properties are that the inverse propagator Δ−1\Delta^{-1} vanishes on-shell and that the potential VV does not have any derivatives. The most general case is

Δ−1=∑k=1∞ak□k,V(ϕ)=∑k=3∞bk⟨ϕk⟩.(24)\Delta^{-1}=\sum_{k=1}^{\infty}a_k\Box^k,\qquad V(\phi)=\sum_{k=3}^{\infty}b_k\langle\phi^k\rangle. \tag*{(24)}

The inverse KLT kernel can be recovered by considering Δ−1=tan⁡(α′π□)\Delta^{-1}=\tan(\alpha'\pi\Box) and V(ϕ)=∑k=3∞Ck−12k+1⟨ϕ2k+1⟩V(\phi)=\sum_{k=3}^{\infty}\frac{C_{k-1}}{2k+1}\langle\phi^{2k+1}\rangle, where CkC_k are Catalan numbers.

Using functional language, the sum rule for (23) reads

M[ϕ]≡ϕδT(0)δϕ−δT(0)δϕϕ+(Δ⋅δT(0)δϕ)δT(0)δϕ−δT(0)δϕ(Δ⋅δT(0)δϕ)=0,(25)\mathcal{M}[\phi]\equiv\phi\frac{\delta T^{(0)}}{\delta\phi}-\frac{\delta T^{(0)}}{\delta\phi}\phi+\left(\Delta\cdot\frac{\delta T^{(0)}}{\delta\phi}\right)\frac{\delta T^{(0)}}{\delta\phi}-\frac{\delta T^{(0)}}{\delta\phi}\left(\Delta\cdot\frac{\delta T^{(0)}}{\delta\phi}\right)=0, \tag*{(25)}

where δT(0)δϕ≡TaδT(0)δϕa\frac{\delta T^{(0)}}{\delta\phi}\equiv T^a\frac{\delta T^{(0)}}{\delta\phi_a} and T(0)T^{(0)} stands for tree-level generator of connected amplitudes. After working through the details, this functional sum rule produces the familiar expression

∑k=2n−1[Δ(X1k)−Δ(Xkn)]A(1…k)A(k…n)=0.(26)\sum_{k=2}^{n-1}\left[\Delta(X_{1k})-\Delta(X_{kn})\right]A(1\ldots k)A(k\ldots n)=0. \tag*{(26)}

We provide the derivation of the above identities in the Appendix.

As an example, let us consider 14Tr⁡ϕ4\frac{1}{4}\operatorname{Tr}\phi^4 theory, where we find

A(123456)−A(234567)=1X14−1X47=(1X14−1X47)A(1234)A(4567).(27)\begin{aligned} A(123456)-A(234567)=\frac{1}{X_{14}}-\frac{1}{X_{47}} \\ &=\left(\frac{1}{X_{14}}-\frac{1}{X_{47}}\right)A(1234)A(4567). \tag*{(27)} \end{aligned}

Derivative Interactions. Let us consider a generalization to theories with derivative interactions. A small modification is required, as the sum rule in the functional language reads

M[ϕ]=[J,δSintδϕ[J]],(28)\mathcal{M}[\phi]=\left[J,\frac{\delta S_{\mathrm{int}}}{\delta\phi}[J]\right], \tag*{(28)}

where J=ϕ+Δ⋅δT(0)δϕJ=\phi+\Delta\cdot\frac{\delta T^{(0)}}{\delta\phi} and M\mathcal{M} is defined by (25). Equation (28) can be extended to virtually any colored theory, including Yang–Mills, since the restriction to spin zero only makes the expressions easier to interpret in terms of amplitudes. For example, the sum rule for Yang–Mills theory in Gervais–Neveu gauge reads:

M[Aμ]=δT(0)δAμ−AμδT(0)δAμ+1□δT(0)δAμ⇌1□δT(0)δAμ=−2ig∂ν([Jμ,Jν]Jμ),(29)\mathcal{M}[A_\mu]=\frac{\delta T^{(0)}}{\delta A_\mu}-A_\mu\frac{\delta T^{(0)}}{\delta A_\mu}+\frac{1}{\Box}\frac{\delta T^{(0)}}{\delta A_\mu}\rightleftharpoons\frac{1}{\Box}\frac{\delta T^{(0)}}{\delta A_\mu}=-2ig\partial^\nu\left([J_\mu,J_\nu]J^\mu\right), \tag*{(29)}

where now Jμ=Aμ−□−1δT(0)δAμJ_\mu=A_\mu-\Box^{-1}\frac{\delta T^{(0)}}{\delta A_\mu}. For scalars we could read these sum rules as statements about amplitudes; here we instead have to understand Mμ=δT(0)δAμM_\mu=\frac{\delta T^{(0)}}{\delta A_\mu} as a Berends–Giele current with one off-shell leg.

String theory

It is natural to ask whether scattering in string theory proper satisfies similar sum rules. We focus on diagonal Z-theory [27], which is given by the integral over the Koba–Nielsen factor [28]

An=∫SL(2,R)dnzz12z23…zn1∏i<j∣zij∣α′sij,(30)A_n=\int_{SL(2,\mathbb{R})}\frac{d^n z}{z_{12}z_{23}\ldots z_{n1}}\prod_{i<j}|z_{ij}|^{\alpha's_{ij}}, \tag*{(30)}

where zij=zi−zjz_{ij}=z_i-z_j. In order to extend our sum rule to these amplitudes, we first rewrite the 1/X1k−1/Xkn1/X_{1k}-1/X_{kn} prefactor as a difference of four-point amplitudes:

A(1…n−1)−A(2…n)=∑k=3n−2A(1…k)A(k…n)×[A(1,k−1,k,k+1)−A(k−1,k,k+1,n)].(31)\begin{aligned} A(1\ldots n-1)-A(2\ldots n)=\sum_{k=3}^{n-2}A(1\ldots k)A(k\ldots n) \\ &\times\left[A(1,k-1,k,k+1)-A(k-1,k,k+1,n)\right]. \tag*{(31)} \end{aligned}

In the context of Z-theory, the four-point amplitudes are Euler–Beta functions A4(i,j,k,l)=B(α′Xik,α′Xjl)A_4(i,j,k,l)=B(\alpha'X_{ik},\alpha'X_{jl}). The rule (31) holds trivially for n=5n=5 and fails for n=6n=6. However, in this case, a closer look reveals that setting s24=0s_{24}=0 recovers the equality. A similar check for n=7n=7 reveals that by setting s24=s25=s35=0s_{24}=s_{25}=s_{35}=0 the desired sum rule is again recovered. These kinematics coincide with minimal kinematics [29, 30, 31, 32, 33, 34, 35, 36]. At this kinematic configuration, the amplitude factorizes completely into Beta functions, or four-point amplitudes. In our case the kinematics is such that all amplitudes in the sum rule factorize in a rather unique way as

A(1…k)=∏j=3k−1A(1,j−1,j,j+1),A(1\ldots k)=\prod_{j=3}^{k-1}A(1,j-1,j,j+1),
A(k…n)=∏j=k+1n−2A(j−1,j,j+1,n).(32)A(k\ldots n)=\prod_{j=k+1}^{n-2}A(j-1,j,j+1,n). \tag*{(32)}

The specific realization of minimal kinematics we are considering can be described by using XX-variables as

X1j,Xjn are free,Xij=∑k=ij−2Xk,k+2 otherwise.(33)X_{1j},X_{jn}\text{ are free},\qquad X_{ij}=\sum_{k=i}^{j-2}X_{k,k+2}\text{ otherwise}. \tag*{(33)}

These minimal kinematics have a neat visual interpretation in the kinematic mesh [37], as depicted in Figure 2. This mesh is a convenient graphical way to organize the planar XijX_{ij}, represented by mesh points, and non-planar sijs_{ij} variables, represented by diamonds. In this figure, the blue and green triangles can be understood as representing the n−1n-1 point amplitudes m(1,2,…,n−1)m(1,2,\ldots,n-1) and m(2,3,…,n)m(2,3,\ldots,n), respectively. Our specific minimal kinematics is the requirement that all sijs_{ij} in the intersection of these two triangles are set to zero.

Figure 2. Kinematic mesh for n=7n=7. In minimal kinematics, the red squares are set to zero. The triangles enclosed by blue and green lines symbolize the two amplitudes A(123456)\mathcal{A}(123456) and A(234567)\mathcal{A}(234567), respectively.

Interestingly, the sum rule (31) becomes a rather trivial statement on this minimal kinematics, where the series telescopes term by term. Using factorizations (32), we might also write an equivalent form2

∑k=2n−1A(1…k+1)A(k…n)−A(1…k)A(k−1…n)=0(34)\sum_{k=2}^{n-1}\mathcal{A}(1\ldots k+1)\mathcal{A}(k\ldots n)-\mathcal{A}(1\ldots k)\mathcal{A}(k-1\ldots n)=0 \tag*{(34)}

which makes the telescoping transparent. That being said, at no point did we have to specify the four-point Z-theory amplitude A4(1234)=B(α′X13,α′X24)\mathcal{A}_4(1234)=B(\alpha' X_{13},\alpha' X_{24}), hence the previous sum rule holds for any fully factorized function (32). It should be noted that the sum rule for Z-theory works only on minimal kinematics, whereas the sum rule holds generally for other theories considered in this work.

Conclusion and Future Directions

In this letter, we have established a novel set of quadratic sum rules that connect color-ordered amplitudes across different multiplicities and, in some cases, across different particle types. Consistency conditions between recursion relations and kinematic shifts gave rise to the first mixed amplitude sum rules, and symbolic regression provides us with the pion complements. The relations among pion and mixed amplitudes can be derived from the sum rule for the inverse diagonal KLT elements via the unity of α′\alpha'-complete scalar amplitudes. We provided a proof of these relations via a functional argument, which further establishes that the same sum rules hold for general scalar theories with non-derivative interactions. This provides a common algebraic identity for a broad class of scattering amplitudes, extending beyond the theories accessible through the unity of scalars.

An open question is whether these relations can be interpreted through the lenses of either surfaceology [32, 38] or positive geometry [39]. Of particular interest are the associahedral grids of [40], which form a common geometric description for the stringy scalar amplitudes. This could help clarify why pion and mixed amplitudes appear together in the sum rules, and how they relate to the analogous statement for their α′\alpha'-completions. It is further natural to consider whether similar relations can be found at loop level. Preliminary investigations suggest a possible extension to both effective theory integrands and the α′\alpha'-complete integrands of [13]. Whether these sum rules hold at all multiplicities and loop orders remains an open problem to which we hope to return in future work.

End Matter

Acknowledgements: We thank Jin Dong, Shruti Paranjape and Jaroslav Trnka for useful discussions and comments on the draft. This work is supported by GAČR 27-27733S, GAUK-340126 and OP JAK CZ.02.01.01/00/22_008/0004632.

End Matter

We present two proofs of the sum rules described in this work. The first proof uses functional methods and applies to a very broad range of theories. The second is substantially simpler, more aligned with standard on-shell methods, but also somewhat weaker.

Functional proof. We first present a derivation of the sum rule (25) through a functional Berends–Giele recursion, the details of which can be found in [13]. Let us consider a theory given by the Lagrangian

L=−12⟨ϕΔ−1ϕ⟩+V(ϕ).(35)\mathcal{L}=-\frac{1}{2}\langle\phi\Delta^{-1}\phi\rangle+V(\phi). \tag*{(35)}

Where ⟨A⟩≡Tr⁡A\langle A\rangle\equiv\operatorname{Tr}A, the inverse propagator is Δ−1=∑k=1∞ak□k\Delta^{-1}=\sum_{k=1}^{\infty}a_k\Box^k and the potential is V(ϕ)=∑k=3∞bk⟨ϕk⟩V(\phi)=\sum_{k=3}^{\infty}b_k\langle\phi^k\rangle. Generator of connected amplitudes T\mathcal{T} is defined by Wick contraction

eiℏT[ϕ]=exp⁡(−iℏ2⟨δδϕ⋅Δ⋅δδϕ⟩)exp⁡(iℏSint[ϕ]).(36)e^{\frac{i}{\hbar}\mathcal{T}[\phi]}=\exp\left(-\frac{i\hbar}{2}\left\langle\frac{\delta}{\delta\phi}\cdot\Delta\cdot\frac{\delta}{\delta\phi}\right\rangle\right)\exp\left(\frac{i}{\hbar}S_{\mathrm{int}}[\phi]\right). \tag*{(36)}

Where dots denote integration over spacetime coordinates ⟨δδϕ⋅Δ⋅δδϕ⟩≡∫dDx dDy δδϕ(x)Δ(x,y)δδϕ(y)\left\langle\frac{\delta}{\delta\phi}\cdot\Delta\cdot\frac{\delta}{\delta\phi}\right\rangle\equiv\int d^Dx\,d^Dy\,\frac{\delta}{\delta\phi(x)}\Delta(x,y)\frac{\delta}{\delta\phi(y)}. Amplitudes are extracted by applying ∫dDx eip⋅xδδϕ(x)\int d^Dx\,e^{ip\cdot x}\frac{\delta}{\delta\phi(x)} on T\mathcal{T} for every particle in the desired scattering process and setting ϕ=0\phi=0. To extract the tree-level generator we expand

around h→0h\to0

T=T0+O(h)(37)\mathcal{T}=\mathcal{T}^{0}+\mathcal{O}(h) \tag*{(37)}

and only consider the leading term T0\mathcal{T}^{0}. The functional form of Berends–Giele recursion takes the form [13]

δT0δϕ=δSintδϕ[J],J=ϕ+Δ⋅δT0δϕ.(38)\frac{\delta\mathcal{T}^{0}}{\delta\phi}=\frac{\delta S_{\mathrm{int}}}{\delta\phi}[J],\qquad J=\phi+\Delta\cdot\frac{\delta\mathcal{T}^{0}}{\delta\phi}. \tag*{(38)}

Where we have used matrix notation

δδϕ≡Taδδϕa,ϕ=ϕaTa.(39)\frac{\delta}{\delta\phi}\equiv T^{a}\frac{\delta}{\delta\phi_{a}},\qquad\phi=\phi^{a}T^{a}. \tag*{(39)}

Since SintS_{\mathrm{int}} does not contain derivatives, we observe that δSintδϕ=∂V(J)∂J\frac{\delta S_{\mathrm{int}}}{\delta\phi}=\frac{\partial V(J)}{\partial J}. Putting everything together the sum rule (25) can be written as

M[ϕ]≡ϕδT0δϕ−δT0δϕϕ+(Δ⋅δT0δϕ)δT0δϕ−δT0δϕ(Δ⋅δT0δϕ)=[ϕ+Δ⋅δT0δϕ,δT0δϕ]=[J,∂V(J)∂J]=0.(40)\mathcal{M}[\phi]\equiv\phi\frac{\delta\mathcal{T}^{0}}{\delta\phi}-\frac{\delta\mathcal{T}^{0}}{\delta\phi}\phi +\left(\Delta\cdot\frac{\delta\mathcal{T}^{0}}{\delta\phi}\right)\frac{\delta\mathcal{T}^{0}}{\delta\phi} -\frac{\delta\mathcal{T}^{0}}{\delta\phi}\left(\Delta\cdot\frac{\delta\mathcal{T}^{0}}{\delta\phi}\right) =\left[\phi+\Delta\cdot\frac{\delta\mathcal{T}^{0}}{\delta\phi},\frac{\delta\mathcal{T}^{0}}{\delta\phi}\right] =\left[J,\frac{\partial V(J)}{\partial J}\right]=0. \tag*{(40)}

Where the first equality is just the definition of a commutator, the second follows from functional Berends–Giele recursion (38) and the commutator vanishes as matrices ∂V(J)∂J\frac{\partial V(J)}{\partial J} and JJ commute.

The functional form of sum rules also allows us to consider theories with derivative interactions. The only new feature is δSintδϕ≠∂V(J)∂J\frac{\delta S_{\mathrm{int}}}{\delta\phi} \ne\frac{\partial V(J)}{\partial J} and we are thus left with a compact expression

M[ϕ]=[J,δSintδϕ[J]],(41)\mathcal{M}[\phi]=\left[J,\frac{\delta S_{\mathrm{int}}}{\delta\phi}[J]\right], \tag*{(41)}

where JJ is defined in (38).

Finally, the amplitude version (1) is obtained by acting n−1n-1-times with ∫ddx eip⋅xδδϕ(x)\int d^d x\,e^{ip\cdot x}\frac{\delta}{\delta\phi(x)} on M\mathcal{M} and color-stripping.

Inductive Proof on Amplitudes. For a general polynomial interaction V(ϕ)=∑p≥3λpϕpV(\phi)=\sum_{p\ge3}\lambda_p\phi^p the Berends–Giele recursion reads

m(1,…,n)=∑p=3nλp∑1=j0<j1<⋯⋯<jp−1=n∏a=1p−1m(ja−1,…,ja)Xja−1ja.(42)m(1,\ldots,n)=\sum_{p=3}^{n}\lambda_p\sum_{\substack{1=j_0<j_1<\cdots\\ \cdots<j_{p-1}=n}}\prod_{a=1}^{p-1}\frac{m(j_{a-1},\ldots,j_a)}{X_{j_{a-1}j_a}}. \tag*{(42)}

We define

Bn:=m(2,…,n)+∑k=3n−2(1X1k−1Xkn)m(1,…,k)m(k,…,n),(43)B_n:=m(2,\ldots,n) +\sum_{k=3}^{n-2}\left(\frac{1}{X_{1k}}-\frac{1}{X_{kn}}\right)m(1,\ldots,k)m(k,\ldots,n), \tag*{(43)}

and show Bn=m(1,…,n−1)B_n=m(1,\ldots,n-1) for all n≥4n\ge4 by induction on nn, which implies (1) for this class of interactions. For n=4n=4 the sum is empty and B4=m(2,3,4)=λ3=m(1,2,3)B_4=m(2,3,4)=\lambda_3=m(1,2,3). We now assume Bp=m(1,…,p−1)B_p=m(1,\ldots,p-1) for 4≤p≤n−14\le p\le n-1.

Step 1: BnB_n does not depend on XinX_{in}, 2≤i≤n−22\le i\le n-2. By (42), every term of BnB_n contains at most one factor 1/Xin1/X_{in} and no other dependence on XinX_{in}. Hence showing that BnB_n has a vanishing residue as Xin=0X_{in}=0 implies that it is independent of the particle index nn. The variable X2nX_{2n} does not appear in (43) at all. For 3≤i≤n−23\le i\le n-2, the pole at Xin=0X_{in}=0 arises from factorization of m(2,…,n)m(2,\ldots,n) and of m(k,…,n)m(k,\ldots,n) with k<ik<i, and from the explicit −1/Xin-1/X_{in} at k=ik=i. This gives

Res⁡Xin=0 Bn=m(i,…,n)[m(2,…,i,n)−m(1,…,i)+∑k=3i−1(1X1k−1Xkn)m(1,…,k)m(k,…,i,n)]=0,(44)\underset{X_{in}=0}{\operatorname{Res}}\,B_n =m(i,\ldots,n)\left[m(2,\ldots,i,n)-m(1,\ldots,i) +\sum_{k=3}^{i-1}\left(\frac{1}{X_{1k}}-\frac{1}{X_{kn}}\right)m(1,\ldots,k)m(k,\ldots,i,n)\right]=0, \tag*{(44)}

where the vanishing follows from the inductive hypothesis.

Step 2: the limit Xin→∞X_{in}\to\infty. Since BnB_n is independent of all XinX_{in}, we may send them to infinity. The terms with 1/Xkn1/X_{kn} drop out, and in (42) for m(k,…,n)m(k,\ldots,n) only the terms with jp−2=n−1j_{p-2}=n-1 survive:

m(k,…,n)→∑pλp∑k=j0<⋯⋯<jp−2=n−1∏a=1p−2m(ja−1,…,ja)Xja−1ja.(45)m(k,\ldots,n)\to\sum_p\lambda_p\sum_{\substack{k=j_0<\cdots\\ \cdots<j_{p-2}=n-1}}\prod_{a=1}^{p-2}\frac{m(j_{a-1},\ldots,j_a)}{X_{j_{a-1}j_a}}. \tag*{(45)}

Inserting (45) into (43) for k=2k=2 and 3≤k≤n−23\le k\le n-2 reproduces (42) for m(1,…,n−1)m(1,\ldots,n-1), with the sum split according to j1j_1: the first term of (43) is j1=2j_1=2 (using m(1,2)/X12=1m(1,2)/X_{12}=1), and the term kk of the sum is j1=kj_1=k. Therefore Bn=m(1,…,n−1)B_n=m(1,\ldots,n-1), which completes the induction.

References

  1. [1]K. Kampf, J. Novotny, and J. Trnka, Phys. Rev. D 87, 081701 (2013), arXiv:1212.5224 [hep-th].DOI
  2. [2]K. Kampf, J. Novotny, and J. Trnka, JHEP 05, 032 (2013), arXiv:1304.3048 [hep-th].DOI
  3. [3]C. Cheung, K. Kampf, J. Novotny, and J. Trnka, Phys. Rev. Lett. 114, 221602 (2015), arXiv:1412.4095 [hep-th].DOI
  4. [4]C. Cheung, K. Kampf, J. Novotny, C.-H. Shen, and J. Trnka, Phys. Rev. Lett. 116, 041601 (2016), arXiv:1509.03309 [hep-th].DOI
  5. [5]F. Cachazo, P. Cha, and S. Mizera, JHEP 06, 170 (2016), arXiv:1604.03893 [hep-th].DOI
  6. [6]C. Bartsch, K. Kampf, J. Novotny, and J. Trnka, Phys. Rev. D 110, 045009 (2024), arXiv:2401.04731 [hep-th].arxiv.org/abs/2401.04731
  7. [7]N. Arkani-Hamed, Q. Cao, J. Dong, C. Figueiredo, and S. He, JHEP 10, 231 (2024), arXiv:2312.16282 [hep-th].
  8. [8]N. Arkani-Hamed, Q. Cao, J. Dong, C. Figueiredo, and S. He, Phys. Rev. D 110, 065018 (2024), arXiv:2401.05483 [hep-th].
  9. [9]J. J. M. Carrasco, C. R. Mafra, and O. Schlotterer, JHEP 06, 093 (2017), arXiv:1608.02569 [hep-th].DOI
  10. [10]H. Kawai, D. C. Lewellen, and S. H. H. Tye, Nucl. Phys. B 269, 1 (1986).
  11. [11]C. Bartsch, K. Kampf, J. Novotný, and J. Trnka, Phys. Rev. D 112, 085002 (2025), arXiv:2505.01501 [hep-th].arxiv.org/abs/2505.01501
  12. [12]F. A. Berends and W. T. Giele, Nucl. Phys. B 306, 759 (1988).
  13. [13]C. Bartsch, K. Kampf, J. Novotný, and J. Trnka, JHEP 10, 099 (2026), arXiv:2602.15102 [hep-th].
  14. [14]R. Kleiss and H. Kuijf, Nucl. Phys. B 312, 616 (1989).
  15. [15]Z. Bern, J. J. M. Carrasco, and H. Johansson, Phys. Rev. D 78, 085011 (2008), arXiv:0805.3993 [hep-ph].DOI
  16. [16]G. Chen and Y.-J. Du, JHEP 01, 061 (2014), arXiv:1311.1133 [hep-th].DOI
  17. [17]J. J. M. Carrasco and L. Rodina, Phys. Rev. D 100, 125007 (2019), arXiv:1908.08033 [hep-th].DOI
  18. [18]K. Kampf, JHEP 12, 140 (2021), arXiv:2109.11574 [hep-th].DOI
  19. [19]T. V. Brown, K. Kampf, U. Oktem, S. Paranjape, and J. Trnka, Phys. Rev. D 108, 105008 (2023), arXiv:2305.05688 [hep-th].arxiv.org/abs/2305.05688
  20. [20]N. E. J. Bjerrum-Bohr, P. H. Damgaard, B. Feng, and T. Sondergaard, Phys. Lett. B 691, 268 (2010), arXiv:1006.3214 [hep-th].
  21. [21]Y. Li, L. Rodina, and Z.-Z. Wang, (2026), arXiv:2111.06810 [hep-th].
  22. [22]S. Mizera, JHEP 06, 084 (2017), arXiv:1610.04230 [hep-th].DOI
  23. [23]S. Mizera, JHEP 08, 097 (2017), arXiv:1706.08527 [hep-th].DOI
  24. [24]M. Crammer, (2023), arXiv:2305.01582 [astro-ph.HI].
  25. [25]S.-M. Urescu and M. Tegmark, Sci. Adv. 6, eaay2631 (2020), arXiv:1905.11481 [physics.comp-ph].
  26. [26]N. Moynihan, (2026), arXiv:2602.15169 [hep-th].arxiv.org/abs/2602.15169
  27. [27]J. Broedel, O. Schlotterer, and S. Stieberger, Fortsch. Phys. 61, 812 (2013), arXiv:1304.7267 [hep-th].DOI
  28. [28]Z. Koba and H. B. Nielsen, Nucl. Phys. B 10, 633 (1969).
  29. [29]C. Cachazo and N. Early, SIGMA 17, 078 (2021), arXiv:2003.07958 [hep-th].DOI
  30. [30]N. Early, JHEP 06, 072 (2024), arXiv:1804.05460 [math.CO].
  31. [31]N. Early, A. Pfister, and B. Sturmfels, (2024), arXiv:2402.03065 [math.AG].arxiv.org/abs/2402.03065
  32. [32]N. Arkani-Hamed, H. Frost, G. Salvatori, P.-G. Plamondon, and H. Thomas, JHEP 09, 033 (2025), arXiv:2311.09284 [hep-th].
  33. [33]Q. Cao, J. Dong, S. He, and C. Shi, Phys. Lett. B 856, 138934 (2024), arXiv:2403.08855 [hep-th].arxiv.org/abs/2403.08855
  34. [34]Q. Cao, J. Dong, S. He, C. Shi, and F. Zhu, JHEP 09, 049 (2024), [Erratum: JHEP 01, 126 (2025)], arXiv:2406.03838 [hep-th].arxiv.org/abs/2406.03838
  35. [35]N. Arkani-Hamed, C. Figueiredo, and G. N. Remmen, JHEP 04, 039 (2025), arXiv:2412.20639 [hep-th].
  36. [36]N. Arkani-Hamed and C. Figueiredo, JHEP 10, 077 (2025), arXiv:2405.09608 [hep-th].
  37. [37]N. Arkani-Hamed, S. He, G. Salvatori, and H. Thomas, JHEP 11, 049 (2022), arXiv:1912.12948 [hep-th].
  38. [38]N. Arkani-Hamed, H. Frost, G. Salvatori, P.-G. Plamondon, and H. Thomas, JHEP 08, 194 (2025), arXiv:2309.15913 [hep-th].
  39. [39]N. Arkani-Hamed, Y. Bai, S. He, and G. Yan, JHEP 05, 096 (2018), arXiv:1711.09102 [hep-th].
  40. [40]C. Bartsch, K. Kampf, D. Podivín, and J. Stalknecht, Phys. Rev. Lett. 136, 011601 (2026), arXiv:2508.20161 [hep-th].DOI

Paper details

vlnc:W_xrep39gv
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents