Lesson 11.5 · 11. Research Frontiers

The Amplituhedron

What is a scattering amplitude?

In particle physics, a scattering amplitude is a number that determines the probability for a set of particles to interact in a certain way. For example, when two electrons meet, the scattering amplitude tells us the probability they will bounce off in a given direction. Computing these amplitudes is the central task of quantum field theory.

What is a Feynman diagram?

A Feynman diagram is a drawing that represents an interaction between particles. Each line represents a particle and each meeting point represents an interaction. To compute a scattering amplitude, one draws all possible diagrams, computes the contribution of each, and adds them up. The more complex the process, the more diagrams must be considered.

What if the most fundamental description of particle interactions does not involve spacetime at all? In 2013, Nima Arkani-Hamed and Jaroslav Trnka discovered a geometric object, the amplituhedron: whose volume directly computes scattering amplitudes without reference to spacetime, virtual particles, or Feynman diagrams. This discovery suggests that spacetime, locality, and even the probabilistic structure of quantum mechanics may be emergent properties of a deeper mathematical structure.

The Core Idea

Scattering amplitudes, the quantities that encode the probabilities for particles to scatter, are not computed by summing Feynman diagrams. Instead, they are the volume of a geometric object called the amplituhedron, which lives in an abstract mathematical space (the Grassmannian) that has no reference to spacetime whatsoever. Locality and unitarity emerge from the geometry.

The amplituhedron: a geometric polytope in the positive Grassmannian whose volume directly computes scattering amplitudes, with no reference to spacetime or Feynman diagrams.

Why Feynman Diagrams Are Not Fundamental

The standard approach to computing scattering amplitudes in quantum field theory uses Feynman diagrams. For a process involving $n$ particles at $L$ loops, one draws all possible diagrams, assigns mathematical expressions to each, and sums them up. This procedure introduces several artifacts:

Virtual particles: Internal lines in Feynman diagrams represent "virtual particles" that are never observed. They are mathematical tools, not physical objects.
Gauge redundancy: Individual diagrams depend on the choice of gauge, even though the physical answer does not.
Enormous cancellations: For complex processes, thousands or millions of diagrams can contribute, with massive cancellations producing a simple final answer. This suggests the diagrams are hiding a simpler structure.

The amplituhedron program reveals what that simpler structure is: a single geometric object whose properties encode the amplitude directly.

The Positive Grassmannian

The mathematical foundation of the amplituhedron is the positive Grassmannian. The Grassmannian $G(k,n)$ is the space of all $k$-dimensional subspaces of an $n$-dimensional vector space, represented as $k \times n$ matrices modulo $GL(k)$. The positive Grassmannian $G_+(k,n)$ is the subset where all ordered maximal minors are positive:

The Positive Grassmannian $$G_+(k,n) = \{ C \in G(k,n) \mid \Delta_{i_1 \cdots i_k}(C) > 0 \text{ for all } 1 \leq i_1 < \cdots < i_k \leq n \}$$ This positivity condition is the mathematical heart of the amplituhedron program. The requirement that all minors be positive is an extremely restrictive condition that defines a beautifully structured region of the Grassmannian.

Defining the Amplituhedron

The tree-level amplituhedron $\mathcal{A}_{n,k}$ is constructed by mapping the positive Grassmannian through external kinematic data. Specifically:

Step 1: External data. The external particles are encoded in momentum twistors $Z_i^A$ ($A = 1, \ldots, k+4$), which are $(k+4)$-dimensional vectors satisfying their own positivity conditions. These automatically incorporate momentum conservation and masslessness.

Step 2: The map. The amplituhedron is the image of the positive Grassmannian $G_+(k,n)$ under the linear map determined by $Z$: $$Y = C \cdot Z^T, \quad C \in G_+(k,n)$$ where $Y$ lives in $G(k, k+4)$. The parameters are: $n$ = number of particles, $k$ = helicity sector (N$^k$MHV), and $k+4$ arises from 4-dimensional spacetime.

Step 3: Loop extension. For $L$-loop amplitudes, each loop adds a 2-plane in $G(2, k+4)$ with additional positivity constraints: $$\mathcal{A}_{n,k,L} \subset G(k, k+4) \times (G(2, k+4))^L$$

Step 4: The canonical form. The amplituhedron possesses a unique canonical form $\Omega_{n,k,L}$: a differential form with logarithmic singularities on all boundaries and no singularities elsewhere. The scattering amplitude is: $$A_{n,k}^{L\text{-loop}} = \int_{\mathcal{A}_{n,k,L}} \Omega_{n,k,L}$$

Emergent Locality and Unitarity

Neither locality nor unitarity appears in the definition of the amplituhedron. Both emerge from its geometric properties:

Spacetime from Geometry

Locality = boundary structure: Physical poles (particles going on-shell) correspond to boundaries where the amplituhedron's geometry degenerates. Different ways of approaching boundaries correspond to different physical channels.
Unitarity = boundary factorization: At each boundary, the amplituhedron factors into a product of lower-point amplituhedra. The residues on poles are products of canonical forms of smaller amplituhedra, this is the geometric content of unitarity.

What is absent from the formulation is remarkable: no spacetime, no Lagrangian, no virtual particles, no gauge redundancy, no Hilbert space. The amplituhedron lives in an abstract Grassmannian, and spacetime appears only when one chooses a particular triangulation (which reproduces the Feynman diagram expansion).

Entanglement and Amplitudes

A rapidly developing research direction treats scattering amplitudes as quantum information objects. The S-matrix acts as a quantum gate on internal degrees of freedom (spin, flavor, color), and its entanglement properties have profound consequences:

Entanglement Minimization and Emergent Symmetry

Requiring the S-matrix to minimally entangle internal degrees of freedom forces the emergence of global symmetries. This principle has been shown to reproduce: Wigner's SU(4) spin-flavor symmetry in nucleon scattering, SO(8) symmetry in the Higgs sector, and SU(N) global symmetry for general S-matrices. Remarkably, imposing entanglement minimization plus low-energy constraints on a general S-matrix ansatz yields an excellent approximation to superstring amplitudes: suggesting that string theory may be the unique theory that minimizes entanglement.

Experimental Evidence

The quantum information content of scattering amplitudes is not merely theoretical. The ATLAS collaboration at the LHC has observed quantum entanglement in top-antitop quark pairs ($t\bar{t}$) at more than $5\sigma$ significance, the highest-energy observation of quantum entanglement ever made, at $\sqrt{s} = 13$ TeV. Bell inequality tests at the LHC are being pursued, though they remain experimentally challenging.

The Surfaceology Revolution (2023-2026)

The biggest development since the original amplituhedron is a complete reformulation of scattering amplitudes in terms of curves on surfaces:

Curve integrals (2023): All-loop scattering amplitudes in Tr($\phi^3$) theory were reformulated as a counting problem on curves drawn on surfaces. No Feynman diagrams, no sum over topologies, a single combinatorial structure.
Unity of theories (2023): A stunning discovery: Tr($\phi^3$) amplitudes secretly contain pion amplitudes (NLSM) and gluon amplitudes (Yang-Mills), related by simple kinematic shifts. One theory, three faces.
Scalar-scaffolded gluons (2024): Yang-Mills theory was "discovered" from elementary counting, scaffolding scalars fuse into gluons, and the binary geometry autonomously guarantees gauge invariance.
Cosmohedra (2024-2026): New geometric objects that compute the wavefunction of the universe, extending the program from flat-space scattering to cosmology.
Kinematic flow (2023): Differential equations governing cosmological correlations derived from pure combinatorics, suggesting that time itself may emerge from mathematical structure.

Two Roads to Emergent Spacetime

Modern theoretical physics has produced two independent programs that both conclude spacetime is not fundamental:

Holography vs. Amplituhedron

Holographic Program	Amplituhedron Program
Spacetime emerges from entanglement	Spacetime absent from scattering
Build spacetime from qubits	Compute observables without spacetime
Gravity = entanglement thermodynamics	Gravity amplitudes = (gauge amplitudes)$^2$
Key tool: tensor networks	Key tool: positive geometry
Works in AdS	Works in flat space

The central open question: are these two programs secretly the same? If so, the connection has not been found. Celestial holography, which maps flat-space amplitudes to a CFT on the celestial sphere, may provide the missing bridge.

Open Questions

The Frontier

Can the amplituhedron program be extended to full QCD and the Standard Model, or is it limited to highly symmetric toy theories?
Is there a "graviton amplituhedron", a geometric object that computes gravity amplitudes directly?
What connects the amplituhedron program to holographic entanglement? No paper has yet connected the canonical form to an entanglement entropy.
Can cosmohedra make predictions for the CMB?
Does the emergence of time from kinematic flow connect to the holographic emergence of spacetime?
Is string theory the unique entanglement-minimizing theory?

Key Insights

The amplituhedron is a geometric object whose volume computes scattering amplitudes in $\mathcal{N}=4$ super Yang-Mills theory, without reference to spacetime, Feynman diagrams, or virtual particles
It is built from the positive Grassmannian, the space of matrices with all positive ordered minors, mapped through external kinematic data encoded in momentum twistors
Locality and unitarity are not assumed but emerge from the boundary structure and factorization properties of the amplituhedron
The surfaceology revolution (2023-2026) has reformulated amplitudes as counting on surfaces, discovered hidden unity between different theories, and extended the program to cosmology
Entanglement minimization of scattering amplitudes forces the emergence of symmetries and may uniquely select string theory
The connection between the amplituhedron program and holographic entanglement remains the great open problem, two roads to emergent spacetime that have not yet met

Key Takeaways

The amplituhedron is a geometric object whose volume directly computes scattering amplitudes without any reference to spacetime, Feynman diagrams, or virtual particles.
Locality and unitarity are not built into the amplituhedron but emerge from its boundary structure and factorization properties.
The surfaceology revolution (2023-2026) reformulated amplitudes as curve-counting on surfaces and revealed hidden unity between seemingly different quantum field theories.
Requiring the S-matrix to minimally entangle internal degrees of freedom forces the emergence of symmetries and may uniquely select string theory.
The amplituhedron program and holographic entanglement represent two independent paths to emergent spacetime whose connection remains the central open question.

← 11.4 Holographic Error Correction 11.6 The Swampland Program →