Curve-Integral Formulation for Tr φ³ Amplitudes

• The paper introduces a curve-integral method that unifies tree- and loop-level amplitude calculations in cubic scalar and deformed string theories.
• It employs scalar scaffolding and kinematical shifts to derive gluon amplitudes, ensuring gauge invariance and reducing combinatorial redundancy.
• The framework maps polytope geometries and tropical methods to traditional Feynman diagrams, streamlining the computation of scattering amplitudes.

A curve-integral (worldsheet-inspired) formulation for stringy Tr $φ^3$ amplitudes provides a uniform, combinatorial, and geometric framework for the computation of both tree-level and loop-level single-trace amplitudes in cubic scalar theories and their stringy deformations. Recent work has established how these constructions encode all field-theory Feynman diagrams, enable powerful reduction algorithms, and directly generalize to produce Yang-Mills, pion, and gluon amplitudes by means of kinematical shifts and differential operators acting on suitably scaffolded scalar variables. This formulation also underpins the planar expansions and the structure of amplitudes in gauge theories, offering canonical mappings between polytope geometries (like the associahedron), worldsheet integrations, and the combinatorics of cubic graphs. Interrelations with tropical geometry, split factorizations, and universal expansions further clarify both field and string limits.

1. Curve-Integral Formulation of Stringy Tr $φ^3$ Amplitudes

The planar single-trace $n$-point amplitude in cubic scalar theory, Tr $φ^3$, admits a “curve-integral” representation akin to an open-string worldsheet formula. In the ‘t Hooft planar limit, the amplitude is given by

$$A_{Tr φ^3}(1,2,…,n) = \int_{Γ} \frac{\prod_{i=1}^n dσ_i}{\mathrm{vol\,}SL(2,ℂ)} \frac{1}{\prod_{α=1}^{n-3} (σ_{a_α} - σ_{b_α})} \exp\left(\sum_{i<j} k_i \cdot k_j \ln|σ_i - σ_j|\right)$$

where:

• The $(a_α, b_α)$ label the chords of a chosen cubic triangulation of the $n$-gon, encoding the planar propagators,
• The SL(2,ℂ) gauge is fixed (e.g., $σ_1 = ∞$, $σ_2 = 1$, $σ_n = 0$),
• The integrand localizes on Feynman-like residue surfaces $σ_{a_α} = σ_{b_α}$,
• Summing residues recovers the sum of color-ordered cubic diagrams: $$A_{Tr φ^3}(1,…,n) = \sum_{g \in \text{Cubic Diagrams}} \frac{1}{\prod_{e \in g} s_e}$$ with $s_e$ the Mandelstam invariants through internal edges (Dong et al., 17 Dec 2025).

This formalism generalizes directly to string theory by including the Koba–Nielsen factor with arbitrary $\alpha'$, connecting to the representation of open-string amplitudes and canonical forms of polytopes, such as the ABHY (Arkani-Hamed–Bai–He–Yan) associahedron in kinematic space (Arkani-Hamed et al., 2019).

2. Kinematic Shifts, Scalar Scaffolding, and Gluons

To construct gluon amplitudes from scalar Tr $φ^3$ curve integrals, an auxiliary “scalar scaffolding” is introduced:

• Each gluon is replaced by a pair of adjacent scalars with modified momenta,
• $k_{2i-1} = k_i + z_i q$, $k_{2i} = k_i - z_i q$, with $q^2 = 0$ and $z_i$ auxiliary,
• The $n$-gluon amplitude becomes the multi-residue under differentiation with respect to these $z_i$,

$$A_{YM}(1,…,n) = \lim_{z_i \to 0} \left[\prod_{i=1}^n \frac{\partial}{\partial z_i}\right]A_{Tr φ^3}^{(2n)}(k_1,…,k_{2n})$$

This procedure “dresses” each gluon as a scalar pair; the derivative operation both fuses the scaffolding and extracts the correct polarization structures (Dong et al., 17 Dec 2025). For bosonic string amplitudes, further even/odd kinematical shifts correspond to replacing $X_{a,b} \to X_{a,b} + \delta^{e,o}_{a,b}$ (with a prescribed sign pattern), producing scaffolded bosonic string integrands (Cao et al., 30 Apr 2025).

3. Differential Operators, Mixed Amplitudes, and Planar Reduction

A striking property of this framework is the existence of $(n-1)$-fold differential operators in the planar variables $X_{i,j}$ (sums of adjacent scalar momenta squared) which, when applied to the $n$-gluon amplitude in its scaffolded scalar form, extract precisely a single planar $φ^3$ diagram: $$D^{(n)}[g]\,A_{YM}(1,…,n) = \frac{1}{\prod_{e \in g} X_e}$$ with $D^{(n)}[g]$ constructed as a product of derivatives $\partial_{X_{i,j}}$ whose arguments correspond to internal chords of the dual diagram (Dong et al., 17 Dec 2025).

More generally, by only converting $r$ scaffolding pairs (leaving mixed amplitudes with $r$ remaining scalars and $(n{-}r)$ gluons), one finds that the space of linearly independent mixed amplitudes is counted by the Catalan number $C_{r-2}$, matching the number of planar $φ^3$ diagrams on $r$ legs. This generalizes the uniqueness of color-ordered gluon amplitudes (unique for $r=0$) to the mixed sector (Dong et al., 17 Dec 2025).

4. Structural Connections: Polytope Geometry, Tropicalization, and Splits

The curve-integral method reveals deep geometric structure:

• The canonical form on the ABHY associahedron, defined via stringy canonical forms, maps directly to the worldsheet curve integrals for $Tr φ^3$ and string amplitudes (Arkani-Hamed et al., 2019),
• The $\alpha'\to 0$ limit recovers the piecewise-linear “tropical” geometry of amplitudes, mapping residues to facets of the polytope and factorization channels (Arkani-Hamed et al., 2019, Arkani-Hamed et al., 2024),
• Split factorizations (in kinematic subspaces, not on poles) manifest naturally as geometrical decompositions of the underlying binary surfaces—amplitudes factor as $$A_n^{(g)}|_{split} = A_{n_1+1}^{(g_1)}A_{n_2+1}^{(g_2)}$$, with recursive structure at all genera (Arkani-Hamed et al., 2024),
• Tropical “headlight functions” and step functions characterize the selection of Feynman diagrams and Schwinger parameters uniformly at any loop order, and the string deformation is implemented as a finite $\alpha'$ replacement of these building blocks (Arkani-Hamed et al., 2024, Arkani-Hamed et al., 2023).

5. Universal Planar Expansions and Nested-Commutator Structure

The universal structure of gluon or mixed amplitudes in Yang-Mills can be reorganized using the curve-integral formalism: $$A_{YM}(1,…,n) = \sum_{m=0}^{n-2} \sum_{|\alpha|=m} (-1)^m [\, \epsilon_1 \cdot f_{\alpha_1}\cdots f_{\alpha_m} \cdot \epsilon_n ]\,A_{YM+φ^3}(\bar{\alpha}\mid1,\alpha,n)$$ where:

• $A_{YM+φ^3}$ are mixed amplitudes with $m$ gluons inserted into skeleton $φ^3$ diagrams on $(n-m)$ scalars,
• Each prefactor $W(1\,g\,n)$ corresponds to a gauge-invariant nested commutator of field strengths in the ordering of the skeleton tree $g$ (Dong et al., 17 Dec 2025),
• This planar expansion dramatically reduces the combinatorial redundancy of reference-ordered expansions, retaining manifest planarity and gauge invariance at every step.

The expansion and its structure constants (nested commutators) also arise as traces over products of field-strength matrices when matching with string-theoretic prefactors under the reduction of open-superstring amplitudes to field theory (Cao et al., 30 Apr 2025).

6. Extensions: All-Loop Structure, General Lagrangians, and Relation to Other Theories

The curve-integral representation is not limited to tree level:

• At arbitrary loop order $L$, Tr $φ^3$ amplitudes decompose into universal $L$-loop “tadpole” kernels and explicit tree-like exponential factors in the kinematic invariants; the curve parameters label the edges of fatgraphs (ribbon graphs), and the dependence on multiplicity $n$ is polynomial in complexity (Arkani-Hamed et al., 2023),
• This structure extends to colored theories, higher valence (non-cubic) interactions, and even to general Lagrangians by employing tropical numerators and combinatorial Wick contractions (Arkani-Hamed et al., 2024),
• Scalar-scaffolded kinematical shifts (including “even–even, odd–odd” mass deformations) unify the curve-integral approach for pions (NLSM) and general gluons, with the same combinatorial mechanism underlying their soft/collinear limits and split properties (Arkani-Hamed et al., 2024),
• Open superstring amplitudes can be algorithmically constructed as sums over shifted stringy Tr $φ^3$ amplitudes, with correction terms for cycles of longer length and carefully chosen linear combinations to cancel tachyonic and higher-derivative spurious terms; the same logic propagates to heterotic and closed-string constructions via the double-copy formalism (Cao et al., 30 Apr 2025).

7. Significance, Applications, and Outlook

The curve-integral formulation for stringy Tr $φ^3$ amplitudes yields profound geometric and combinatorial simplifications:

• Encodes all field-theory and string-theory information in canonical integrals over moduli spaces, sidestepping the redundancies of Feynman diagrammatics,
• Provides the underlying mechanism for the CHY (Cachazo–He–Yuan) formulation, double-copy relations, and universal expansions in gauge theories (He et al., 2018),
• Connects positive geometry, polytopes, and tropical mathematics directly to physical kinematics, clarifying the interplay between locality, unitarity, and algebraic factorization,
• Enables systematic all-order recursive and analytical reductions of amplitudes, including in the presence of gauge or scalar–gluon mixing, at tree and loop level,
• Allows for efficient enumeration, factorization, and computation of amplitudes in Yang-Mills, non-linear sigma models, and (super)string theory, with prospects for future generalizations to arbitrary valence, color structures, and effective actions.

These developments establish the curve-integral approach as a unifying principle at the intersection of scattering amplitudes, combinatorics, and geometry, providing a platform for ongoing advances in mathematical physics and quantum field theory (Dong et al., 17 Dec 2025, Arkani-Hamed et al., 2019, Arkani-Hamed et al., 2024, Cao et al., 30 Apr 2025, Arkani-Hamed et al., 2023, Arkani-Hamed et al., 2024, He et al., 2018).