A Bound on Holographic Entanglement Entropy from Inverse Mean Curvature Flow

TL;DR

This paper establishes an upper bound on holographic entanglement entropy in (2+1)-dimensional CFTs using inverse mean curvature flow, linking geometric and energetic properties of the dual AdS spacetime.

Contribution

It introduces a generalized inverse mean curvature flow approach to derive a new bound on entanglement entropy in holographic CFTs, akin to a Penrose inequality.

Findings

The entanglement entropy is bounded by a weighted local energy density.

The bound is analogous to a Penrose inequality in AdS spacetimes.

The result holds perturbatively in 1/N and is conjectured to be more generally valid.

Abstract

Entanglement entropies are notoriously difficult to compute. Large-N strongly-coupled holographic CFTs are an important exception, where the AdS/CFT dictionary gives the entanglement entropy of a CFT region in terms of the area of an extremal bulk surface anchored to the AdS boundary. Using this prescription, we show -- for quite general states of (2+1)-dimensional such CFTs -- that the renormalized entanglement entropy of any region of the CFT is bounded from above by a weighted local energy density. The key ingredient in this construction is the inverse mean curvature (IMC) flow, which we suitably generalize to flows of surfaces anchored to the AdS boundary. Our bound can then be thought of as a "subregion" Penrose inequality in asymptotically locally AdS spacetimes, similar to the Penrose inequalities obtained from IMC flows in asymptotically flat spacetimes. Combining the result with positivity of relative entropy, we argue that our bound is valid perturbatively in 1/N, and conjecture that a restricted version of it holds in any CFT.