Talks

I suggest a minimal practical formal structure for a more fundamental theory than the Standard Model + GR and review a mechanism that produces such a structure.  The proposed mechanism has possibilities of producing non-canonical phenomena in SU(2) and SU(3) gauge theories which might allow conditional predictions that can be tested.

The slides and other writings are posted on my web page

     http://www.physics.rutgers.edu/~friedan/#Perimeter

During my visit to PI, I hope also to discuss informally a separate project in pure QFT, a scheme to construct a new kind of QFT of extended objects (also described on my web page).

[joint work with: Victor Albert, John Preskill (Caltech), Sepehr Nezami, Grant Salton, Patrick Hayden (Stanford University), and Fernando Pastawski (Freie Universität Berlin)]

Quantum error correction and symmetries are concepts that are relevant in many physical systems, such as in condensed matter physics and in holographic quantum gravity, and play a central role in error correction of reference frames [Hayden et al., arXiv:1709.04471]. I will show that codes that are covariant with respect to a continuous local symmetry necessarily have a limit to their ability to serve as approximate error-correcting codes against erasures at known locations. This is because the environment necessarily gets information about the global logical charge of the encoded state due to the covariance of the code. Our bound vanishes either in the limit of large individual subsystems, or in the limit of a large number of subsystems; in either case there exist codes which approximately achieve the scaling of our bound and become good covariant error-correcting codes. Our results can be interpreted as an approximate version of the Eastin-Knill theorem, quantifying to which extent it is not possible to carry out universal computation approximately on encoded states. In the context of holographic AdS/CFT, our approach provides some insight on time evolution in the bulk versus time evolution on the boundary. We expect further implications for symmetries in holography and in condensed matter physics.

I will present a brief introduction to non-Lorentzian geometries, an important example of such geometries being Newton-Cartan geometry and its torsionful generalization, which is the natural geometry to which non-relativistic field theories couple to. The talk will subsequently review how such geometries have in recent years appeared in gravity, string theory and holography. In particular, torsional Newton-Cartan geometry has been shown to appear as the boundary geometry for Lifshitz spacetimes. Furthermore, dynamical Newton-Cartan geometry is related to Horava-Lifsthiz gravity theories and appears in novel Chern-Simons theories of gravity in three dimensions. The latter can be obtained from a well-defined limit of the AdS3/CFT2 correspondence. Finally, I will briefly comment on how Newton-Cartan geometry appears in non-relativistic string theory.

The Bondi mass loss formula has been central in the context of early research on gravitational waves. We show how it can be understood as a particular case of BMS current algebra and discuss the associated central extension. We then move down to three dimensions where a more complete picture emerges.

Suppose the eigenvalue distributions of two matrices $M_1$ and $M_2$ are known. What is the eigenvalue distribution of the sum $M_1+M_2$? This problem has a rich pure mathematics history dating back to H. Weyl (1912) with many applications in various fields. Free probability theory (FPT) answers this question under certain conditions, which often involves some degree of randomness (disorder). We will describe FPT and show examples of its powers for the qualitative understanding (often approximations) of physical quantities such as density of states, and gapped vs. gapless phases of some Floquet systems. These physical quantities are often hard to compute exactly. Nevertheless, using FPT and other ideas from random matrix theory excellent approximations can be obtained. Besides the applications presented, we believe the techniques will find new applications in new contexts.

There has been a dissatisfaction with the postulates of quantum mechanics essentially since the moment that those postulates were first written down. Over the years since there have therefore been many attempts (some successful and some less so) to reconstruct quantum theory from various sets of postulates. The aim being to gain a deeper understanding of the theory by providing a conceptually clear underpinning from which the standard formalism can be derived. In this talk I present recent work with Carlo Maria Scandolo and Bob Coecke (arXiv:1802.00367) in which we reconstruct quantum theory from purely diagrammatic postulates within the process theory framework, showing that the conceptual bare-bones of quantum theory concerns the manner in which systems and processes compose.

Generically, a small amount of matter introduced to anti-de Sitter spacetime leads to formation of a black hole; however, the high degree of symmetry of AdS means that some initial distributions of matter (possibly also technically generic) oscillate indefinitely. Whether a given initial profile leads to a horizon at arbitrarily small amplitudes is of great interest for a number of reasons, not least because horizon formation corresponds holographically to thermalization in CFT.  We will present an overview of approaches the question and show a phase diagram of the stability behavior of AdS in the presence of a scalar field, including an analysis of nonperturbative phases at the border of perturbative stability and instability.

The next frontiers in cosmic microwave background (CMB) science include a detailed mapping of the CMB polarization anisotropy, with goals of detecting the inflationary B-mode signal and reconstructing high-fidelity maps of the matter distribution via CMB lensing, as well as a first detection of CMB spectral distortions.  At this level of precision (~nK), Galactic and extragalactic foregrounds may be the ultimate limiting factor in deriving cosmological constraints.  I will discuss biases due to foregrounds in CMB lensing measurements, including the first calculation of the lensing bias due to the kinematic Sunyaev-Zel’dovich effect, as well as recent progress in developing novel foreground-free CMB lensing estimators.  I will then present methods to extend CMB foreground parameterizations in a systematic, flexible way, with applications to both polarization and spectral distortion measurements.  Using this framework, I will discuss spectral distortion forecasts for CMB spectrometer mission concepts, showing that high-significance measurements of the Compton-y and relativistic thermal Sunyaev-Zel’dovich signals can be expected, as well as a potential detection of the primordial mu-type distortion due to Silk damping of small-scale acoustic modes.

Over the last decade, there have been enormous gains in machine learning technology primarily driven by neural networks. A major reason neural networks have outperformed older techniques is that the cost of optimizing them scales well with the size of the training dataset. But neural networks have the drawback that they are not very well understood theoretically.

Recent work by several groups has explored an alternative approach to creating machine learning model functions based on tensor networks, which are a technique for parameterizing many-body quantum wavefunctions. The cost of training tensor network models scales similarly to the cost of training neural networks. In addition, their relatively simple, linear structure has provided good theoretical understanding of their properties, and underpins many powerful techniques to optimize and manipulate them. 

After introducing tensor network machine learning models, I will discuss some of the techniques to optimize them and results for supervised and generative machine learning tasks. These techniques can automatically adapt the number of parameters and suggest interesting interpretations and extensions for 'deep' tensor network variants. I will conclude by discussing a recent tensor network based proposal to formulate hybrid quantum-classical algorithms for machine learning with quantum computers.