---
title: 'Amplituhedron: Geometry of Scattering Amplitudes'
url: https://www.emergentmind.com/topics/amplituhedron
type: topic
---

# Amplituhedron: Geometry of Scattering Amplitudes

The amplituhedron is a geometric object that encodes scattering amplitudes in planar $\mathcal{N}=4$ supersymmetric Yang–Mills (SYM) theory as canonical differential forms on a positive region in a Grassmannian. The central principle is that geometric positivity and stratification in Grassmannian spaces subsume the analytic properties—locality, unitarity, and dual/conformal invariance—traditionally imposed in quantum field theory. Formally, the amplituhedron generalizes cyclic polytopes and the positive Grassmannian to a class of semialgebraic regions whose canonical volume forms reproduce, via explicit pushforward, the full integrand for (tree- and loop-level) scattering in planar $\mathcal{N}=4$ SYM [1312.2007]. In what follows, the combinatorics, topology, stratification, convexity, and physical consequences of the amplituhedron are rigorously detailed.

## 1. Formal Definition and Construction

Let $G_+(k,n)$ denote the positive Grassmannian—the set of $k$-planes in $\mathbb{R}^n$ represented by $k\times n$ real matrices $C$ with all maximal $k\times k$ minors strictly positive. External data is given by an $(k+m)\times n$ matrix $Z$ whose maximal minors are strictly positive. The amplituhedron is the image
\[
A_{n,k,m}(Z) = \{ Y = C\cdot Z \in G(k,k+m) \mid C \in G_+(k,n) \}
\]
where $G(k,k+m)$ is the Grassmannian of $k$-planes in $\mathbb{R}^{k+m}$ [1312.2007]. In the case $m=4$, this map is central to planar $\mathcal{N}=4$ SYM.

For the extension to loop level with $L$ loops, one considers a family of $D_{(i)} \in G_+(2,n)$ (one for each loop), stacking them with $C$ into a $(k+2L) \times n$ matrix $\mathcal{C}$, and forming $\mathcal{Y} = \mathcal{C} \cdot Z$. Positivity must be imposed for all maximal minors of $\mathcal{C}$, including mixed minors involving different $D_{(i)}$ and (optionally) $C$ rows [1408.3410].

In the $m=1$ case, the amplituhedron is equivalent to the complex of bounded faces for a cyclic hyperplane arrangement, and for $m=2$ it arises as a nonlinear analog of cyclic polytopes in $G(k,k+2)$ [1608.08288, 2501.08221]. For $k=1$, $A_{n,1,m}(Z)$ recovers the classical cyclic polytope [1312.2007].

## 2. Canonical Form and Physical Correspondence

For any positive geometry $(X,X_{\geq 0})$, the canonical form $\Omega(X_{\geq 0})$ is the unique (up to scale) top-degree rational differential form with logarithmic poles on all boundary divisors, such that the residue along any complete flag of boundaries yields $\pm1$ [1312.2007].

The integrand of planar $\mathcal{N}=4$ SYM is this canonical form on $A_{n,k,m}(Z)$:
\[
\Omega(A_{n,k,m}) = \bigwedge_{a} d\log f_a(x,Z)
\]
in local positive coordinates $x_a$. The rational denominators $f_a$ are affine-linear functions whose vanishing defines the boundary facets of the amplituhedron [1408.3410].

Alternatively, in bracket notation (e.g., for $n=4$, $k=0$, $L=2$),
\[
\Omega_{2\text{-loop}} = 
\frac{ \langle AB\ 34 \rangle \langle CD\ 12 \rangle + \langle AB\ 23 \rangle \langle CD\ 14 \rangle + \dots }
     { \langle AB\ CD \rangle \prod_{i<j\in\{1,2,3,4\}} \langle AB\ ij \rangle \langle CD\ ij \rangle }
\, d\mu_{AB} d\mu_{CD}
\]
with each bracket a $2\times2$ or $4\times4$ minor [1408.3410].

The poles of $\Omega$ correspond exactly to co-dimension one boundaries of the amplituhedron, implemented by the vanishing of minors or bracket expressions, and with the numerator canceling would-be spurious poles incompatible with positivity [1408.3410, 1312.7878].

## 3. Stratification and Boundary Structure

Every boundary of the amplituhedron is realized by sending a collection of Plücker minors to zero. The stratification is naturally two-staged:

- The initial decomposition ("$\Gamma_0$") is by positroid cells: determining which $2\times2$ (or general $k\times k$) minors vanish, as in the positroid stratification [1408.3410]. For loops, this is the $L$-fold (product) structure, restricted by extended positivity.
- The secondary stratification ("$\Gamma_1$") is by vanishing non-minimal minors, subject to compatibility with Plücker relations and extended positivity.

Boundary labels record which minors are set to zero and determine the dimension. For example, in the four-point, $k=0$, two-loop case:
\[
\text{Dimension } d: \quad N^{(d)}\ \text{boundaries}
\]
with $N^{(d)} = \{1,8,36,104,178,224,216,128,34\}$ for $d=8$ down to $d=0$ respectively, totaling 1232 boundaries. The corresponding Euler characteristic is
\[
\chi = \sum_{d=0}^{8} (-1)^d N^{(d)} = 2
\]
[1408.3410].

Permutation labels provide an alternative combinatorial stratification corresponding to decorated permutations for positroid cells at tree level, and to more intricate combinatorial data for multi-loop amplituhedra.

## 4. Deformations, Topology, and Explicit Examples

A remarkable discovery is that, after formally relaxing all inter-minor constraints in stratification—specifically, allowing all non-minimal minors to be independently switched off—the "deformed amplituhedron" possesses a simplified topology: for $n=4$, $k=0$, the Euler characteristic for the deformed object is always 2 for $L\geq2$ loops. This suggests a conjectured uniform topological structure for the deformed positive Grassmannian in these cases, though the geometric reason remains elusive [1408.3410].

Explicit tabulations for two- and three-loop cases at low $n$ are given, with the Euler characteristic switching from $-14$ (undeformed) to $2$ (deformed) at three loops, emphasizing the impact of stratification combinatorics on topology.

## 5. Amplituhedron–Integrand Correspondence

The structure of the canonical form is tightly controlled by the amplituhedron stratification. The poles of the integrand match the boundary strata ("labels") of the amplituhedron: each denominator corresponds to a minor vanishing and, through extended positivity and Plücker relations, allows systematic enumeration of all codimension strata both from the geometric and analytic sides [1408.3410].

This correspondence provides nontrivial tests of the amplituhedron/scattering amplitude duality, as the full list of boundaries and their multiplicities extracted from the integrand coincides with that derived geometrically via minors.

## 6. Alternative Descriptions and Theoretical Generalizations

The combinatorial and topological structure of the amplituhedron admits alternative characterizations:

- **Sign variation and cyclic permutations:** In the $m=1$ amplituhedron, sign variation characterizes the cell decomposition and the topology (PL homeomorph to a closed ball), realized as the complex

Source: https://www.emergentmind.com/topics/amplituhedron