Search Talks

The last decade has seen a wave of characterizations of quantum theory using the formalism of generalized probability theory.

In this talk, I will introduce a novel operational approach to characterizing and reconstructing quantum theory which puts an observer’s information acquisition -- rather than the probability structure – centre stage. In particular, we consider an observer interrogating a system with binary questions and explain how an elementary set of rules governing the observer’s acquisition of information about the system leads to qubit quantum theory. The derivation is constructive, elucidating, among other things, the origin of entanglement, monogamy and more generally the correlation structure. This approach also yields a new characterization of pure states in terms of ‘conserved informational charges’ which, in turn, define the unitary group.

In this talk I will propose a general correspondence which associates a non-perturbative quantum mechanical operator to a toric Calabi-Yau manifold, and I will propose a conjectural expression for its spectral determinant. As a consequence of these results, I will derive an exact quantization condition for the operator spectrum. I will give a concrete illustration of this conjecture by focusing on the example of local P2. This approach also provides a non-perturbative Fermi gas picture of topological strings on toric background and suggests the existence of an underlying theory of M2 branes behind this formulation.

With the increase of the center-of-mass energy from 8 TeV
to 13 TeV for LHC Run 2, the probability for boosted topologies will
become even higher than in Run 1. This also comes with a large
increase in pileup from the increased luminosity. This talk
investigates the state of the art of boosted algorithms and grooming
techniques, addresses shortcomings and possible improvements, and
discusses hot-topic items that will be interesting early on in Run 2.

A renormalization group transformation implements a scale transformation while resetting the UV cut-off, so that the theories before and after the RG transformation contain the same degrees of freedom, but with a modified action. In Wilson's original proposal, the cut-off is reset by integrating out the degrees of freedom between the old and new cut-off. In recent years we have learned that, alternatively, one can decouple these degrees of freedom by means of local unitary transformations (disentanglers). The result is a reversible renormalization group flow, and an efficient description of many-body ground states. In this talk, I will show you pretty drawings and will use fancy names, such as "multi-scale entanglement renormalization ansatz" (MERA). MERA is currently studied as a possible lattice realization of the AdS/CFT correspondence. I will keep it simple.

After 50 years of dreaming about it, space-based microlensing observations are now underway.  A 2014 100-hr Spitzer Pilot Program generated "microlens parallaxes" for dozens of lenses, opening the prospect of measuring the Galactic distribution of planets.  This program will be expanded 8-fold in 2015.  Analogous observations by Kepler will measure the mass function of free-floating planets.

WFIRST microlensing observations will, as advertised, "complete the planetary census" but they will do an immense amount of astrophysics as well.  I discuss how microlensing's take off builds on rapid, ongoing, ground-based developments.

Does a generic quantum system necessarily thermalize? Recent developments in disordered many-body quantum systems have provided crucial insights into this long-standing question. It has been found that sufficiently disordered systems may fail to thermalize leading to a 'many-body localized' phase. In this phase, the fundamental assumption underlying equilibrium statistical mechanics, namely, the equal likelihood for all states at the same energy, breaks down. A fundamental question is: what happens as the disorder becomes weaker so that one approaches the localization-delocalization transition? For example, does the system thermalize *at* the transition? 

In this talk, I will show that very general considerations on the scaling of entanglement entropy close to the transition imply that at a continuous many-body localization transition, the system is necessarily ergodic.

Finally, I will present recent results on "eigenstate thermalization", a long standing hypothesis which posits that a single eigenstate hides within itself a thermal ensemble. In particular, I will discuss which class of operators do or do not satisfy eigenstate thermalization.

Using holography, I will describe an approach for understanding the physics of a big bang singularity by translating the problem into the language of the dual quantum field theory. Certain two-point correlators in the dual field theory are sensitive to near-singularity physics in a dramatic way, and this provides an avenue for investigating how strong quantum gravity effects in string theory might modify the classical description of the big bang.

I will talk about the physics of models in which dark matter consists of composite bound states carrying a large conserved dark “nucleon” number. The properties of sufficiently large dark nuclei may obey simple scaling laws, and this scaling can determine the number distribution of nuclei resulting from Big Bang Dark Nucleosynthesis. For plausible models of asymmetric dark matter, dark nuclei of large nucleon number, e.g. >~ 10^8, may be synthesised, with the number distribution taking one of two characteristic forms, which interestingly are broadly independent of initial conditions. A possible consequence of these scenarios is alterations to direct detection signals, which may be coherently enhanced (relative to collider signals), and could be modified by new momentum-dependent form factors. Inelastic interactions between dark matter states might also be important in astrophysical settings.