Non-saturation of Bootstrap Bounds by Hyperbolic Orbifolds

TL;DR

This paper investigates whether bounds from the conformal bootstrap are truly saturated by hyperbolic orbifolds, revealing they are not, and derives formulas for their OPE coefficients using modular forms.

Contribution

It demonstrates that bootstrap bounds are not saturated by hyperbolic orbifolds and introduces formulas for their OPE coefficients based on modular form links.

Findings

Bootstrap bounds are not saturated by hyperbolic orbifolds.

Explicit formulas for OPE coefficients of hyperbolic orbifolds.

Bounds closely approximate actual values but do not exactly match.

Abstract

In recent years the conformal bootstrap has produced surprisingly tight bounds on many non-perturbative CFTs. It is an open question whether such bounds are indeed saturated by these CFTs. A toy version of this question appears in a recent application of the conformal bootstrap to hyperbolic orbifolds, where one finds bounds on Laplace eigenvalues that are exceptionally close to saturation by explicit orbifolds. In some instances, the bounds agree with the actual values to 11 significant digits. In this work we show, under reasonable assumptions about the convergence of numerics, that these bounds are not in fact saturated. In doing so, we find formulas for the OPE coefficients of hyperbolic orbifolds, using links between them and the Rankin-Cohen brackets of modular forms.