New Approaches to Scattering AmplitudesNew Approaches to Scattering Amplitudes From On-Shell...
Transcript of New Approaches to Scattering AmplitudesNew Approaches to Scattering Amplitudes From On-Shell...
New Approaches to Scattering Amplitudes
From On-Shell Diagrams to the Amplituhedron
Based on: arXiv:1207.0807, S. Franco arXiv:1408.3410, S. Franco, D. Galloni, A. Mario6, J. Trnka arXiv:1502.02034, S. Franco, D. Galloni, B. Penante, C. Wen
Sebastián Franco
The City College of New York
Seba
stia
n Fr
anco
2
Outline
� Introduction and Motivation
� On-shell diagrams
� From on-shell diagrams to the Grassmannian
v Canonical variables
v Towards a combinatorial classification of general on-shell diagrams
v Non-planar boundary measurement
� The Amplituhedron
� Conclusions
� A systematic approach to non-planar on-shell diagrams
Seba
stia
n Fr
anco
3
Introduction and Motivation
� The standard approach to it is based on Feynman diagrams, which make locality and unitarity manifest
� Our description of the microscopic world is based on Quantum Field Theory
� Not all symmetries are manifest in the Feynman Diagram approach ⇒ inefficient
� For concreteness, let us focus on N=4 SYM:
Ø Expectation: some of the lessons can be extrapolated to “more standard” QFTs
Ø Very symmetric example of a 4d interacting QFT
Towards a New Formulation of QFT
� Goals: v Study a new formulation of QFT that makes all such symmetries manifest
v Initiate a systematic investigation of the non-planar case, developing the technology that is necessary for doing so
Seba
stia
n Fr
anco
4
� The new formulation is based on on-shell diagrams Arkani-‐Hamed, Bourjaily, Cachazo, Goncharov, Postnikov, Trnka
� The elementary building blocks are the 3-point MHV and MHV super-amplitudes
� In the planar limit, scattering amplitudes can be constructed in terms of planar on-shell diagrams
On-Shell Diagrams
1 2
3 4
1
4
2
3
1 2
3 4
5
Which can be glued using on-shell lines to give rise to more complicated diagrams
Seba
stia
n Fr
anco
5
Points in G(k,n):
Ø Rows: n-dimensional vectors spanning the planes
n = # scattered particles k = # negative helicity
The Grassmannian G(k, n) : space of k-dimensional planes in n dimensions
On-Shell Diagrams and the Grassmannian
C = ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
n
k Ø Up to a GL(k) redundancy
1) Pick a perfect orientation:
Franco, Galloni, MarioM (multiple boundaries) Franco, Galloni, Penante, Wen (arbitrary diagrams)
� On-shell diagrams provide a powerful parametrization of the Grassmanian Postnikov (disk) Gekhtman, Shapiro, Vainshtein (annulus)
From On-Shell Diagrams to G(k,n)
Seba
stia
n Fr
anco
6
8
1
5
4
7 6
2) The entries in the matrix C are given by the boundary measurement:
1 2 3 4 5 6 7 8 1 1 C12 C13 0 C15 0 0 C18
4 0 C42 C43 1 C45 0 0 C48
6 0 C62 C63 0 C65 1 0 C68 7 0 C72 C73 0 C75 0 1 C78
C = Example:
Plücker coordinates = k × k minors
� Every on-shell diagram is associated to a differential form:
The On-Shell Form
on-shell form Ω
Arkani-‐Hamed, Bourjaily, Cachazo, Goncharov, Postnikov, Trnka
2 3
Seba
stia
n Fr
anco
7
� There is an infinite number of on-shell diagrams
Moves and Reductions
� A pressing quesVon: find global diagnostics for equivalence and reductions that apply to non-planar graphs
Equivalence
Bubble Reduction
Merger Square move (every diagram can be made bipartite)
� They can be organized into equivalence classes and connected by reductions
More generally, deleting and edge while covering the same region of the Grassmannian
Seba
stia
n Fr
anco
8
� It is possible to introduce GL(1)-invariant generalized face variables
Generalized face variables
� Automatically dlog form:
Embedding the Diagram
� It is useful to consider the embedding of the diagram into a bordered Riemann surface
For F faces, B boundaries and genus g:
v Faces: they are subject to ∏ f i =1
v Cuts: B-1 paths connecting pairs of boundaries
v Fundamental cycles: αm and βm pairs, m=1,…,g
d = #edges - #nodes = F + B + 2g - 2 = F – χembed.
Volume Form
Degrees of Freedom
Franco, Galloni, Penante, Wen
Seba
stia
n Fr
anco
9
Seba
stia
n Fr
anco
10
v Plücker matching: when two on-shell diagrams give rise to the same sets of Plücker coordinates and relations
A General Classification of On-Shell Diagrams
v Reduction: an on-shell diagram B is a reduction of an on-shell diagram A, if it is obtained from A by deleting edges and it covers the same region of the Grassmannian
v Reduced graph: it is impossible to remove edges while covering the same region of the Grassmannian
� How do we organize the infinite number of on-shell diagrams? It is desirable to develop a classification of generic on-shell diagrams
Franco, Galloni, Penante, Wen (arbitrary) Arkani-‐Hamed, Bourjaily, Cachazo, Goncharov, Postnikov, Trnka (planar)
the only thing we care about for leading singularities
� For planar diagrams, it can be combinatorially phrased in terms of the properties of zig-zag paths and permutations
General Definitions
not enough for non-planar diagrams
Ø Two necessary conditions for equivalence are Plücker matching and having the same number of degrees of freedom
Seba
stia
n Fr
anco
11
A Useful Tool: Perfect Matchings
� Perfect matchings play a central role in connecting bipartite on-shell diagrams to combinatorics and geometry
� (Almost) Perfect Matching: p is a subset of the edges in the graph such that:
Ø Every internal node is the endpoint of exactly on edge in p
Ø Every external point belongs to either one or zero edges in p
� Finding the perfect matchings reduces to calculating the determinant of an adjacency matrix of the graph (Kasteleyn matrix), and some generalizations Franco
p1 p2 p3 p4
p5 p6 p7
Seba
stia
n Fr
anco
12
Polytopes from On-Shell Diagrams Franco, Galloni, Penante, Wen
1. Generalized Matching Polytope
Piµ = 1 if Xi ∈ pµ
0 if Xi ∈ pµ
Xi = ∏ pµ P iµ
µ
� Let us define the (#edges × #perfect matchings)-dimensional matrix P:
� The following map between edge weights and perfect matchings automatically satisfies all relations between edge weights:
Xi : edge pµ : perfect matching
� Generalized Matching Polytope: there is a distinct point in the polytope for every perfect matching pµ, with a position vector given by the corresponding column in the matrix P
� We can associate two types of polytopes to every on-shell diagram
Useful for?
Seba
stia
n Fr
anco
13
1 2
3 4
¤ 4 degrees of freedom ¤ 7 perfect matchings 7 points
4d polytope
A Simple Example: the Top-‐Cell of G(2,4)
Linear relations between positions of points
Relations between edge weights
� The dimensionality of the generalized matching polytope is equal to the number of degrees of freedom in the diagram
Seba
stia
n Fr
anco
14
2
� Projection of the matching polytope that only preserves information on the external legs
Perfect matchings which coincide on all
external legs
Plücker coordinate
Generalized Matroid Polytope
� Multiple perfect matchings can be projected to the same point in this polytope
Useful for?
1 2
3 4
Example: the Top-‐Cell of G(2,4)
� Every point in the generalized matching polytope corresponds to a Plücker coordinate
� All relations among them are encoded in the positions of the points
Seba
stia
n Fr
anco
Combinatorial Classification
v Plücker matching: two on-shell diagrams give rise to the same sets of Plücker coordinates and relations if their generalized matroid polytopes coincide
v Reduction: an on-shell diagram B is a reduction of an on-shell diagram A, if it is obtained from A by deleting edges and it has the same generalized matroid polytope of A
v Reduced graph: it is impossible to remove edges while preserving the generalized matroid polytope
� The classification of general on-shell diagrams we introduced above has a powerful implementation in terms of generalized matching and matroid polytopes
Franco, Galloni, Penante, Wen
� It is interesting to compare this global approach with one for planar graphs based on permutations. For example, permutations generically change under reduction
� A novel feature of non-planar diagrams: new constraints between Plucker coordinates beyond Plucker relations can arise Systematically captured by the polytopes
Franco, Galloni, Penante, Wen Arkani-‐Hamed, Bourjaily, Cachazo, Postnikov, Trnka
Seba
stia
n Fr
anco
Genus-1 Examples
� None of these diagrams admits a genus-0 embedding
1 2
3
4 5
6
1
2 3
4
5 6
� Both diagrams have the same number of degrees of freedom
1 0 0 0 0 0 0 1 1 0 1 1 0 1 1 1 1 0 0 1 1 2 0 1 0 1 0 1 0 1 0 1 0 0 0 1 0 1 0 1 1 1 3 0 1 0 1 1 0 0 1 1 0 0 0 0 1 1 0 1 1 0 1 4 0 0 1 1 1 0 0 0 0 0 1 0 1 1 1 0 1 0 1 1 5 0 0 1 1 0 1 0 0 0 0 1 1 0 1 0 1 1 1 1 0 6 0 0 0 0 0 0 1 1 1 0 1 1 0 1 1 1 1 1 0 0 6 2 2 1 2 2 2 1 2 2 1 2 2 1 1 1 1 1 1 1
G1 =
1 0 0 0 0 0 0 1 1 0 1 1 0 1 1 1 1 0 0 1 1 2 0 1 0 1 0 1 0 1 0 1 0 0 0 1 0 1 0 1 1 1 3 0 1 0 1 1 0 0 1 1 0 0 0 0 1 1 0 1 1 0 1 4 0 0 1 1 1 0 0 0 0 0 1 0 1 1 1 0 1 0 1 1 5 0 0 1 1 0 1 0 0 0 0 1 1 0 1 0 1 1 1 1 0 6 0 0 0 0 0 0 1 1 1 0 1 1 0 1 1 1 1 1 0 0 5 3 3 2 3 3 1 1 2 2 1 2 2 2 1 1 2 2 2 2
G2 =
Only the multiplicities change
Generalized Matroid Polytopes
(1) (2)
in fact, they are equivalent
Seba
stia
n Fr
anco
17
The Non-Planar Boundary Measurement
� Order external nodes using boundaries and cuts (for B>1)
Ø Consistency with the combinatorial classification
� ObjecVves:
Ø The general expression must reduce to the one for planar graphs (positivity) Postnikov
Higher genus do this within unit cell
A Proposal
1 3
Franco, Galloni, Penante, Wen
� Close flows into loops by using boundaries and cuts
4 5 1. Overall sign for Cij entry: (-1)
n ij
nij = number of sources between i and j
2. Additional sign for each contribution: (-1)r+1 r = rotation number = clock - counterclock
Plücker coordinates are sums of contributions from all perfect matchings at the corresponding point in the generalized matroid polytope
2
Seba
stia
n Fr
anco
18
An Example
� A B=2, g=1 example:
Seba
stia
n Fr
anco
Beyond Plücker Relations
� Non-planar diagrams can exhibit a new type of pole, at which no Plücker coordinate vanishes
� However, it is possible to remove an edge while preserving the matroid polytope (i.e. the non-vanishing Plücker coordinates) a relation is generated
Franco, Galloni, Penante, Wen Arkani-‐Hamed, Bourjaily, Cachazo, Postnikov, Trnka
� The relations are systematically accounted for by the matching and matroid polytopes, when expressing Plücker coordinates in terms of perfect matchings
� The dimension of the graph is equal to 9, i.e. equal to the one of G(3,6)
Combinatorial Manifestation
19
1
2 4
3
5 6
A G(3,6) example:
Seba
stia
n Fr
anco
20
Seba
stia
n Fr
anco
21
� The amplituhedron is a new algebraic geometric construction that maximally geometrizes scattering amplitudes in planar N=4 SYM in terms of a single object:
The Amplituhedron
v Amplitude = Volume
v Different representations of the amplitude from different “triangulations”
Arkani-‐Hamed, Trnka
Beyond the Positive Grassmannian
L(1)
⋮ L(L)
Y
D(1)
⋮ D(L)
C
= · Z Loops
Tree-level
Amplituhedron
Z ∈ M+(4+k,n) C ∈ G+(k,n) D(i) ∈ G(2,n)
k = 0 : MHV
(Z) D(i)
C
D(i)
D(j)
C (C) , , , ,
⋯
� Extended positivity:
Seba
stia
n Fr
anco
22
� Ideas that are similar to the ones presented in this talk can used for stratifying the amplituhedron
Galloni, Franco, MarioM, Trnka
Anatomy of the Amplituhedron
� Precisely these 1232 boundaries can alternatively be determined from the integrand of the scattering amplitude, providing an exquisite test of the amplituhedron/scattering amplitude correspondence
1 9 44 140 274 330 264 136 34
Example: 2 loops, 4 particles, k=0 χ = 2
Seba
stia
n Fr
anco
Conclusions
� We initiated a systematic study of non-planar on-shell diagrams in N=4 SYM
Ø Canonical variables
Ø Combinatorial classification (i.e. region matching and reductions) in terms of generalized matching and matroid polytopes
� We carried out an extensive study of the geometry of the amplituhedron
Ø Boundary measurement for completely general on-shell diagrams
23
� We developed powerful technology to do so, including:
Ø We stratified the 4-particle MHV amplituhedra for L=2 and 3, finding remarkably simple topologies and exact agreement with the singularity structure of the amplitude. We also developed a combinatorial implementation of it.
Ø We introduced and studied a class of deformations of the amplituhedron. For 4-particle MHV, which have χ=2 at least up to L=4
Ø It might be possible to exploit these simple deformed geometries to constraint the integrand
The Future Se
bast
ian
Fran
co
� There seems to exist an intriguing relation, yet to be explored, with recently studied deformations of the Grassmannian
On-Shell Diagrams and Quivers
1
3 4
2 5
6
9
8
7
10
9
1
3 4
2
6
8
7
10
� Interesting connection between on-shell diagrams and a general class of quiver gauge theories, known as bipartite field theories (BFTs)
Ferro, Lukowski, Meneghelli, Ple]a, Staudacher Beisert, Broedel, Rosso
Franco Yamazaki, Xie
� Classify non-equivalent reduced diagrams for top-dimensional cells of G(k,n). How can they be generated? Bourjaily, Franco, Galloni, Wen (in progress)
25 Seba
stia
n Fr
anco