Connecting Archimedean and non-Archimedean AdS/CFT

Abstract

This thesis develops a non-Archimedean analog of the usual Archimedean anti-de Sitter (AdS)/conformal field theory (CFT) correspondence. AdS space gets replaced by a Bruhat-Tits tree, which is a regular graph with no cycles. The boundary of the Bruhat-Tits tree is described by an unramified extension of the p-adic numbers, which replaces the real valued Euclidean vector space on which the CFT lives. Conformal transformations on the boundary act as linear fractional transformations.

In the first part of the thesis, correlation functions are computed in the simple case of massive, interacting scalars in the bulk. They are found to be surprisingly similar to standard holographic correlation functions down to precise numerical coefficients, when expressed in terms of local zeta functions. Along the way, we show that like in the Archimedean case, CFT conformal blocks are dual to geodesic bulk diagrams, which are bulk exchange diagrams with the bulk points of integration restricted to certain geodesics. Other than these intriguing similarities, significant simplifications also arise. Notably, all derivatives disappear from the operator product expansion, and the conformal block decomposition of the four-point function. Finally, a minimal bulk action is constructed on the Bruhat-Tits tree for a single scalar field with nearest neighbor interactions, which reproduces the two-, three-, and four-point functions of the free O(N) model.

In the second part, the p-adic O(N) model is studied at the interacting fixed point. Leading order results for the anomalous dimensions of low dimension operators are obtained in two separate regimes: the epsilon-expansion and the large N limit. Remarkably, formulae for anomalous dimensions in the large N limit are valid equally for Archimedean and non-Archimedean field theories, when expressed in terms of local zeta functions. Finally, higher derivative versions of the O(N) model in the Archimedean case are considered, where the general formula for anomalous dimensions obtained earlier is still valid. Analogies with two-derivative theories hint at the existence of some interesting new field theories in four real Euclidean dimensions.