Quantum Information

In a theory of gravity, observables must be diffeomorphism-invariant. Such observables are nonlocal, in contrast with the usual formulation of local quantum field theory. Working to leading order in Newtons constant G, I'll describe a construction of diffeomorphism-invariant observables for a scalar field coupled to gravity that closely parallels an analogous construction for charged particles in electrodynamics. These observables acting on the vacuum create scalar particles together with their (linearized) gravitational dressing. The commutator of two such spacelike-separated observables is nonvanishing at order G, and is related to the gravitational potential between two masses. Based on arXiv:1507.07921 with Steve Giddings.

In recent years, tensor networks have been proposed as a useful framework for understanding holographic duality, especially the relation between quantum entanglement and space-time geometry. Most tensor networks studied so far are defined in the large scale compared with AdS radius. In this talk, I will describe a new tensor network approach which defines a holographic mapping that applies to a refined network with sub-AdS scale resolution, or even to a flat space. The idea of quantum error correction code plays an essential role in this approach. Using this new tensor network, we can study features of the bulk theory, such as how locality at sub-AdS scale emerges in a "low energy subspace" even though the whole theory is intrinsically nonlocal, as a quantum gravity theory should be.

If entanglement entropy in a small geodesic ball is maximized at fixed volume in the vacuum, then it should be stationary under variation to a nearby state. I will show that this stationarity condition is equivalent to the semiclassical Einstein equation. If the matter QFT is not conformal, then the derivation requires a further assumption about QFT, whose validity is currently under investigation. [Based on http://arxiv.org/abs/1505.04753]

We discuss area terms in entanglement entropy and show that a recent formula by Rosenhaus and Smolkin is equivalent to the Adler-Zee sum rule for the renormalization of the Newton constant in terms of correlator of traces of the stress tensor. We elaborate on how to fix the ambiguities in these formulas: Improving terms for the stress tensor of free fields, boundary terms in the modular Hamiltonian, and contact terms in the Euclidean correlation functions. We make computations for free fields and show how to apply these calculations to understand some results for interacting theories which have been studied in the literature. We check the sum rule holographicaly. We also discuss an application to the F-theorem.

The fact that a black hole is a fast-scrambler is at the heart of black hole information paradoxes. It has been suggested that chaos can be diagnosed by using an out-of-time correlation function, which is closely related to the commutator of operators separated in time. In this talk I propose that the tripartite information (also known as topological entanglement entropy) can be used as a quantitative information theoretic measure of chaos. By viewing a quantum channel as a state via the Choi-Jamilkowski isomorphism, the tripartite information measures four-party entanglement between the “past” and the “future”, much like an out-of-time correlation function. I will compute the time-evolution of the tripartite information for three systems; (a) non-integrable spin systems on a lattice, (b) planar networks of perfect tensors which mimic the growth of the Einstein-Rosen bridge and (c) a holographic system. This talk is based on an ongoing work with Xiaoliang Qi and Daniel Roberts.

We investigate causality constraints on the time evolution of entanglement entropy after a global quench. We analyze three models for the spread of entanglement: holographic quenches, the free particle streaming model of Calabrese and Cardy generalized to higher dimensions and an arbitrary pattern of entanglement, and a particle model with an infinite scattering rate. In these models we exhibit the intricate interplay of causality, strong subadditivity, and maximally entangled subsystems. Finally, using strong subadditivity we prove that the normalized rate of growth of entanglement entropy is bounded by the speed of light.