Princeton University Gravity Initiative Seminar

Abstract: In AdS/CFT, a spacetime geometry is said to contain a python's lunch when there exist choices of boundary regions with associated entanglement wedges that contain locally but not globally minimal surfaces. Previously, such geometries have been connected to black hole evaporation and modelled with tensor networks featuring ancillas both added and projected, which has in turn led researchers to conjecture that reconstructing information from past the locally minimal surface is computationally difficult. In this work, we use quantum cryptographic tools related to a primitive known as the Conditional Disclosure of Secrets (CDS) to further develop consequences of the projective tensor network model. We argue from the tensor network picture that the mutual information between appropriate CFT subregions is lower bounded linearly by an area difference associated with the geometry of the lunch, which has bulk geometric consequences we can check. We prove weakened versions of this geometrical statement in asymptotically AdS3 spacetimes satisfying the null energy condition, and confirm it in some example geometries.