Intertwining Operator Realization of the AdS/CFT Correspondence

TL;DR

This paper provides a group-theoretic framework for the AdS/CFT correspondence by constructing intertwining operators that relate bulk AdS fields to boundary conformal fields, generalizing previous approaches.

Contribution

It introduces a general setting for the AdS/CFT correspondence using representation theory and constructs boundary-to-bulk operators as intertwining operators, extending prior work.

Findings

Bulk fields are shown to have two boundary counterparts, emphasizing their equal footing.

The framework generalizes previous models by using representations of the Euclidean conformal group.

Boundary-to-bulk operators are explicitly constructed as intertwining operators.

Abstract

We give a group-theoretic interpretation of the AdS/CFT correspondence as relation of representation equivalence between representations of the conformal group describing the bulk AdS fields $\phi$ and the coupled boundary fields $\phi_0$ and ${\cal O}$. We use two kinds of equivalences. The first kind is equivalence between bulk fields and boundary fields and is established here. The second kind is the equivalence between coupled boundary fields. Operators realizing the first kind of equivalence for special cases were given by Witten and others - here they are constructed in a more general setting from the requirement that they are intertwining operators. The intertwining operators realizing the second kind of equivalence are provided by the standard conformal two-point functions. Using both equivalences we find that the bulk field has in fact two boundary fields, namely, the coupled boundary fields. Thus, from the viewpoint of the bulk-boundary correspondence the coupled fields are on an equal footing. Our setting is more general since our bulk fields are described by representations of the Euclidean conformal group $G=SO(d+1,1)$, induced from representations $\tau$ of the maximal compact subgroup $SO(d+1)$ of $G$. From these large reducible representations we can single out representations which are equivalent to conformal boundary representations labelled by the conformal weight and by arbitrary representations $\mu$ of the Euclidean Lorentz group $M=SO(d)$, such that $\mu$ is contained in the restriction of $\tau$ to $M$. Thus, our boundary-to-bulk operators can be compared with those in the literature only when for a fixed $\mu$ we consider a 'minimal' representation $\tau=\tau(\mu)$ containing $\mu$.