Loops and legs: ABJM amplitudes from $f$-graphs

This paper initiates a systematic method to reconstruct planar loop integrands of ABJM scattering amplitudes from a symmetric generating function of squared amplitudes based on $f$-graphs, demonstrating that individual amplitudes can be disentangled for various multiplicities and loop orders up to six loops, thereby establishing a structural parallel with $\mathcal{N}=4$ super Yang-Mills theory.

The Gist

Imagine you are trying to solve a massive, complex puzzle. In the world of theoretical physics, this puzzle is understanding how subatomic particles scatter (bounce off each other) in a specific universe called ABJM theory.

For a long time, physicists have tried to calculate these scattering events loop-by-loop, like trying to build a skyscraper one brick at a time. It's slow, tedious, and prone to errors.

This paper introduces a revolutionary new way of looking at the problem. Instead of building the skyscraper brick by brick, the authors suggest looking at a photograph of the finished building and working backward to figure out exactly how every single brick was placed.

Here is the breakdown of their discovery using simple analogies:

1. The "Shadow" vs. The "Object"

In physics, calculating the exact path of a particle (the Amplitude) is incredibly hard. However, calculating the probability of a collision happening (the Squared Amplitude or "Cross-section") is often easier.

Think of the Amplitude as a complex 3D sculpture.
Think of the Squared Amplitude as the shadow that sculpture casts on the wall.

Usually, if you only see the shadow, you can't tell exactly what the 3D object looks like. The shadow loses information. But the authors discovered something magical: In this specific theory (ABJM), the "shadow" is so detailed and structured that it actually contains all the information needed to reconstruct the original 3D sculpture perfectly.

2. The "f-graph" Blueprint

The authors use a tool called an f-graph. Imagine these as a special kind of flowchart or a network of dots and lines.

• The Dots: Represent points in space and time.
• The Lines: Represent the connections and forces between them.

In the past, physicists had to draw these graphs for every single specific scenario (e.g., "4 particles colliding at 2 loops"). It was a mess of different drawings.

The paper reveals that there is a Master Blueprint (called $F_N$). This single blueprint is a giant, symmetrical network that includes every possible scenario at once. It has a hidden "super-symmetry," meaning if you shuffle the dots around, the blueprint looks the same. It's like a kaleidoscope where the pattern remains perfect no matter how you twist it.

3. The "Disentanglement" Trick

The big challenge is: How do you pull one specific story out of this giant, mixed-up blueprint?

The authors developed a method to "disentangle" the specific answers from the general mix.

• The Analogy: Imagine a smoothie made of strawberries, bananas, and blueberries. Usually, once blended, you can't get the fruit back. But the authors found a special "strainer" (mathematical tools called Yangian invariants and soft cuts) that can separate the strawberry flavor from the banana flavor, even though they were blended together in the "shadow."

They showed that by using these strainers, they can extract the exact details of:

• 4-particle collisions (up to 6 loops deep).
• 6-particle collisions (up to 2 loops deep).
• 8-particle collisions (at the tree level).

4. The "Odd" Problem

There was a weird rule in this theory: You can't have an "odd" number of loops in the squared shadow. It's like saying you can only see the shadow of a building if the sun is at an even-numbered hour.

The authors proved that even though the "shadow" (the squared amplitude) disappears for odd loops, the Master Blueprint still holds the secret. They found a way to use the even-loop shadows to mathematically deduce what the odd-loop collisions look like. It's like deducing what a building looks like at midnight by studying its appearance at noon and 2 PM.

5. Why This Matters

This isn't just about solving one specific puzzle. It suggests a deep, hidden order in the universe.

• Unification: It shows that "loops" (complexity) and "legs" (number of particles) are two sides of the same coin.
• Efficiency: Instead of calculating millions of complex equations for every new scenario, physicists might just need to analyze this one "Master Blueprint" and use their "strainers" to get the answer instantly.
• New Geometry: The paper hints that the universe might be built on a strange, hidden geometry (related to "positive Grassmannians" and "amplituhedrons") that we are just beginning to understand.

Summary

The authors of this paper found a universal "Master Blueprint" for particle collisions. They proved that even though this blueprint looks like a jumbled mess of probabilities (the squared amplitude), it actually contains the complete, step-by-step instructions for every single collision event. They invented a new mathematical "decoder ring" to extract these instructions, showing that the universe is far more interconnected and symmetrical than we previously thought.

It's a bit like realizing that the entire Library of Congress is hidden inside a single, perfectly folded piece of origami, and they just figured out how to unfold it.

Technical Summary

1. Problem Statement

The paper addresses a fundamental challenge in the study of scattering amplitudes in the three-dimensional $\mathcal{N}=6$ superconformal Chern-Simons-matter theory (ABJM theory).

• Context: In planar $\mathcal{N}=4$ Super Yang-Mills (SYM) theory, a powerful framework exists where all-loop integrands are encoded in a permutation-invariant generating function, $F_N$, constructed from "f-graphs" (planar graphs with specific weights). This function unifies amplitudes with different numbers of external legs ($n$) and loop orders ($L$) under the constraint $N = n + L$.
• The Gap: A similar generating function, $F_N$, was recently proposed for ABJM theory. However, extracting individual scattering amplitudes from this squared object in ABJM is significantly more difficult than in SYM due to:
  1. Parity Constraints: ABJM squared amplitudes vanish for odd loop orders and odd particle numbers, making the extraction of odd-loop integrands non-trivial.
  2. Graphical Complexity: Unlike SYM, where $n$-point $L$-loop integrands are simply identified by $n$-faces in $F_{n+L}$, the ABJM case involves parity-odd structures (involving $\epsilon$-tensors) and "kissing" topologies that obscure the direct graphical extraction.
  3. Dimensional Constraints: In $D=3$, Gram determinant identities reduce the independence of certain f-graphs, introducing free parameters that must be fixed.

Core Question: Can the individual planar integrands of ABJM amplitudes (for arbitrary $n$ and $L$) be systematically reconstructed from the permutation-invariant squared amplitude $F_N$?

2. Methodology

The authors employ a multi-faceted approach combining graphical analysis, algebraic constraints, and physical consistency conditions:

• f-Graph Analysis: They utilize the weight-3 bipartite f-graphs that constitute $F_N$. These graphs manifest the hidden $S_N$ permutation symmetry. Solid lines represent poles ($1/x_{ij}^2$) and dashed lines represent numerators ($x_{ij}^2$).
• Disentangling Products: To extract the $L$-loop integrand $A^{(L)}_n$ from the squared object $M^{(L)}_n$, the authors must subtract products of lower-loop amplitudes (e.g., $A^{(\ell)} A^{(L-\ell)}$).
  • For 4-point amplitudes, they identify specific sub-topologies (box-wheels) within the planar f-graphs that correspond to products of lower-loop amplitudes. By isolating these, they recursively extract the pure $L$-loop integrand.
  • They relate the squared amplitude to the logarithm of the amplitude ($\log R_4$), utilizing "negative geometries" (bipartite pole structures) to organize the result.
• Ansatz Construction with Constraints: For higher multiplicities ($n=6, 8$), they construct an ansatz for the amplitude based on:
  • Yangian Invariants: Using BCFW building blocks and leading singularities.
  • Dual Conformal Invariance (DCI): Ensuring the integrand has the correct conformal weight.
  • Symmetry: Imposing cyclicity, reflection symmetry, and parity properties.
  • Soft-Cut Conditions: Using the soft limit ($x^2_{i,i+1} \to 0$) as a recursion relation to fix free parameters in the ansatz.
• Leading Singularity Squaring: They derive a determinant formula for squaring super-functions in the orthogonal Grassmannian $OG_+(k, 2k)$. This reveals a systematic vanishing pattern for products of leading singularities depending on the parity of $k$ (the number of particles/2).

3. Key Contributions and Results

A. Four-Point Amplitudes ($n=4$)

• Extraction Mechanism: The authors demonstrate that the $L$-loop integrand cannot be found by a simple face-counting rule. Instead, they show that odd-loop integrands (which are parity-odd) can be extracted by identifying "box-wheel" subgraphs in the planar f-graphs.
• Recursive Reconstruction: They provide a recursive procedure to obtain the $L$-loop planar integrand from $F_{4+L}$.
• New Results:
  • They provide a new planar representation for the 5-loop 4-point integrand (previously known only in bipartite/non-planar forms).
  • They compute the 6-loop 4-point integrand and the logarithm of the amplitude up to $L=6$.
  • They establish a connection between the bipartite f-graphs and the "negative geometries" of the ABJM amplituhedron, fixing coefficients for specific topologies (e.g., the coefficient of the $K_{3, L+1}$ star graph is $2^{L-1}$).

B. Six-Point Amplitudes ($n=6$)

• Tree-Level: They derive the tree-level amplitude by squaring the sum of leading singularities from the top cell of the orthogonal Grassmannian. They determine the overall sign by matching to the known squared amplitude $M^{(0)}_6$.
• Loop Extraction:
  • They construct an ansatz for the 1-loop and 2-loop integrands involving parity-even triangles and parity-odd boxes.
  • By matching the squared ansatz against the f-graph data and applying soft-cut conditions, they resolve a 2-dimensional ambiguity in the coefficients.
  • Result: They successfully extract the 1-loop integrand and the parity-even part of the 2-loop integrand, confirming that the squared amplitude contains sufficient information to reconstruct the individual amplitudes.

C. Eight-Point Amplitudes ($n=8$)

• Tree-Level Extraction: They extend the method to 8 points, where the tree amplitude involves co-dimension 1 cells of the Grassmannian.
• Vanishing Pattern Discovery: A crucial theoretical finding is the proof (Theorem 4.3) that the product of leading singularities from different branches/solutions vanishes systematically based on the parity of $k$ (where $n=2k$):
  • If $k$ is even: Products between different branches vanish.
  • If $k$ is odd: Products between the same branches vanish.
• Result: This allows them to isolate the correct tree-level amplitude by squaring the ansatz and matching it to the f-graph limit, reproducing known results and confirming the validity of the extraction method for higher multiplicities.

4. Significance

• Unification of Loops and Legs: The paper provides strong evidence that the "loops and legs" unification framework, successful in $\mathcal{N}=4$ SYM, holds true for ABJM theory. The squared amplitude $F_N$ is a complete repository of information for all individual amplitudes.
• New Computational Tools: The development of techniques to disentangle products of lower-loop amplitudes from f-graphs (specifically the box-wheel identification and soft-cut recursion) provides a new toolkit for calculating high-loop amplitudes in 3D theories.
• Connection to Geometry: The work deepens the link between scattering amplitudes and positive geometry (the ABJM amplituhedron) by showing how bipartite f-graphs encode the pole structures of the logarithm of the amplitude.
• Future Directions: The results open the door to calculating higher-loop cusp anomalous dimensions and exploring the "squared amplituhedron" and "correlahedron" in 3D via dimensional reduction, potentially shedding light on the amplitude/Wilson loop/correlator triality in ABJM theory.

In summary, the paper successfully bridges the gap between the abstract, permutation-symmetric squared amplitudes and the concrete, physical scattering integrands in ABJM theory, establishing a systematic extraction method that parallels and extends the success of f-graphs in $\mathcal{N}=4$ SYM.