An exercise in Color-Dual Cut Tiling: $\mathcal{N}=8$ Supergravity from Positivity

TL;DR

This paper introduces a new color-dual cut tiling method that simplifies multi-loop amplitude calculations in supersymmetric theories by leveraging tree-level BCJ relations and positivity constraints, reducing the need for explicit cuts.

Contribution

It presents an agglomerative algorithm for constructing integrands from minimal on-shell data, enabling the derivation of non-planar amplitudes from planar cuts without explicit loop-level BCJ relations.

Findings

Sufficient planar cut information can generate non-planar sectors.

Tree-level BCJ relations reduce the number of necessary cuts.

Method applies to three-loop two-to-two scattering in maximally supersymmetric theories.

Abstract

The BCJ duality between color and kinematics brings two advantages to calculating multi-loop scattering amplitudes. First the number of ordered cuts that need to be performed to fix the integrand to a gauge theory is minimal -- reducing the factorially-growing number of diagrams to a single-digit basis in all known cases. Second is the trivial generation of related gravitational amplitudes from gauge theory amplitudes via double-copy. Mounting evidence suggests there are some cases where no local color-dual representations exist, even when the semi-classical theory is color-dual. Can we still simplify the calculations without making the duality manifest at the level of the integrand? Here we take a non-trivial step in this direction by showing that the satisfaction of tree-level BCJ relations is sufficient to dramatically reduce the number of explicit cuts that must be performed, even when the loop-level relations are not explicitly satisfied. We introduce an agglomerative algorithm, color-dual cut tiling, that identifies and builds the entire integrand from the simplest on-shell conditions applied to a seed of off-shell integrand information. Specifically, we demonstrate that for two-to-two scattering at three loops in the maximally supersymmetric gauge theory there is sufficient information contained in planar cuts -- completely determined by positivity constraints -- to generate all of the non-planar sector. Additionally, we make use of the generalized double copy to generate a representation of maximally supersymmetric gravity as a functional of the planar SYM input. We discuss how the process might generalize, and then close by commenting on the applicability of this method for additional cases of interest where performing explicit unitarity cuts is expensive.