The Setting
Take the points and the faces , each face a triple of points. Listed from face to face , each with its points in increasing order, the faces are (1)✦ The graph is their incidence graph: its vertex set has vertices, and each face is joined to its three points. A corner of is a pair of a face and a slot .✦ The point of a corner is the entry in slot of face in the table (1).✦ The corners are the edges of , one for each incidence between a point and a face.
The question is whether satisfies Sidorenko’s conjecture: whether for every graphon .
Finite Laws and the Bipartite Moment
A finite probability law on a finite set is a function with for every and .✦ The expectation of a function under is (2)✦
The product law of finite probability laws on , one for each in a finite set , is the finite probability law on the maps whose weight at is .✦ When every is , the product law is written .
Let and be finite probability laws on finite sets and , and let be a bipartite kernel. The kernel mean of is (3)✦ The bipartite moment of , the homomorphism density of in with the points of sent to and its faces to , is (4)
The Claim
Theorem 1. There is a symmetric counter kernel✦: a nonempty finite set , a finite probability law on and a function with for all , whose kernel mean is positive and whose bipartite moment satisfies (5)✦ For every finite probability law on a finite set and every function , drawing all vertices of independently from gives the bipartite moment: (6)✦ where assigns a point of to each vertex of , to the point and to the face .
In the paper, the result reads: assuming Lemma 3.1, there is a symmetric kernel on a finite probability space with mean such that , where stands for the left side of (6) and for .
What Is Checked
The claims (5) and (6) are stated over symmetric counter kernels, over finite probability laws on finite sets with their expectations (2) and product laws, and over the kernel mean (3) and the bipartite moment (4) of the incidence graph of the faces (1).