The application of artificial neural network to central questions in the theory of quantum matter is a rapidly developing field. The insight driving the field is that the problems of theoretical interests are primarily those of regression in which an exponentially large volume of data must be condensed into a more accessible or meaningful form, e.g., labeled with phases. In this talk, I will review the state of this rapidly developing field. I will then showcase how my group built and taught neural networks to recognize topological phases, different non-equilibrium phases and universal features from scanning tunneling microscopy data. I will then discuss new insights from the synergy between human intelligence and artificial intelligence: the key to our success.
The realization of large-scale controlled quantum systems is an exciting frontier in modern physical science. In this talk, I will introduce a new approach based on cold atoms in arrays of optical tweezers. We use atom-by-atom assembly to deterministically prepare arrays of individually controlled cold atoms. A measurement and feedback procedure eliminates the entropy associated with the probabilistic trap loading and results in defect-free arrays of over 60 atoms . Strong, coherent interactions are enabled by coupling to atomic Rydberg states. We realize a programmable Ising-type quantum spin model with tunable interactions and system sizes of up to 51 qubits. Within this model we observe transitions into ordered states (Rydberg crystals) that break various discrete symmetries, verify high-fidelity preparation of ordered states, and investigate dynamics across the phase transition in large arrays of atoms .
An alternative, hybrid approach for engineering interactions is the coupling of atoms to nanophotonic structures in which guided photons mediate interactions between atoms. I will discuss our progress towards entangling two atoms that are coupled to a photonic crystal cavity and I will outline the exciting prospects of this approach for scaling the system to large distances in a quantum network.
I will discuss recent progress at extending the conformal bootstrap to 4-point functions containing the stress-energy tensor. This progress includes deriving analytical sum rules for the coefficients in stress-tensor 3-point functions and producing numerical bounds on these coefficients in 3D CFTs for various gaps in the spectrum. In the latter analysis we obtain the first determination of the stress-tensor 3-point function and parity-odd gap in the critical 3D Ising model.
For over two decades, polarized He-3 has proven to be a powerful tool for investigating the structure of the neutron. The advent of liter-scale polarized He-3 targets, in turn, led quickly to magnetic resonance imaging of the gas space of human lungs with unprecedented resolution. Clinical research using "noble-gas imaging" quickly came to include the use of polarized Xe-129, and both gases have seen extensive use in research related to pulmonary disease and drug discovery. Because xenon dissolves readily into blood, there has also been interest in imaging other organ systems. Outside the lungs, however, results have been limited by small signals. In addition to providing some historical context, I will describe a new imaging modality that builds on noble-gas imaging, but that has enormously increased sensitivity. Spatial information is encoded using magnetic-field gradients, but imaging data are acquired entirely through the detection of gamma rays (without the use of a gamma camera). The quantity of atoms needed for producing images is reduced by more than a factor of a million, creating the potential for a new type of nuclear tracer. Indeed, so few atoms are required that novel polarization techniques can be considered as a path for expanding the range of applications.
About the speaker
Professor Gordon Cates conducts research in three diverse areas spanning atomic, nuclear, and medical physics. Unifying these activities is the use of optical pumping and spin exchange, techniques that make it possible to polarize the spins of electrons, atoms and nuclei using light sources such as lasers. Critical to such research is the study of spin interactions during atomic collisions, spin-relaxation at surfaces, and numerous aspects of laser physics.
A major thrust of Prof. Cates’ research has been understanding the “spin-structure” of the neutron. These efforts, involving electron scattering from spin-exchange polarized 3He targets at SLAC and Jefferson Laboratory (JLab), have helped shed light on the foundations of QCD and the question of what it is within the nucleon that carries spin. Prof. Cates has also been involved in studying parity violation in polarized electrons scattering in order to search for the presence of strange quarks in the nucleon (at JLab) and for physics beyond the standard model (at SLAC).
Another important aspect of Prof. Cates’ research involves a new type of MRI in which a subject inhales a laser-polarized noble gas such as 3He or 129Xe which is subsequently imaged. Coinvented by Prof. Cates in the early 1990’s, “hyperpolarized gas imaging” produces images of the gas space of the lungs of unprecedented resolution. Also, since Xe dissolves readily into the blood and various tissues in the body, there is potential to extend the technique to other organs. Many basic physics issues remain critical to the continuing development of hyperpolarized gas imaging.
Precision measurements of large, coherent ensembles of aligned fermionic spins are powerful probes of many fundamental physics question, such as tests of Lorentz violation, one of the pillars of general relativity; searches for new sources of CP-violation to explain the baryon asymmetry of the universe; and probes of high energy symmetries via their relic goldstone bosons. In this talk I will cover some recent developments in experimental techniques that have greatly improved the sensitivity of these measurements, and avenues for further improvements. I will also discuss some new ideas about ultra-low-mass axions and fuzzy dark matter candidates, and the prospects for using polarized-spin systems as dark matter detectors.
Lorenzo's research is in theoretical high-energy astrophysics. Lorenzo investigate the origin of non-thermal emission from Pulsar Wind Nebulae (PWNe), AGN jets, gamma-ray bursts (GRBs), supernovae, galaxy clusters, and low-luminosity accretion flows like Sgr A* at the center of our Galaxy.
It is still a mystery how these objects can accelerate particles up to the highly non-thermal energies required to explain the observed spectra, that typically extend from the radio up to the gamma-ray band.
By means of ab initio large-scale plasma simulations, Lorenzo investigate particle acceleration in shocks and magnetic reconnection from first principles, with the aim of using the simulations to interpret the observations, and ultimately unveil the nature of astrophysical non-thermal sources.
The sub-GeV dark matter mass range has received increased interest in the last several years, owing to the lack of any unambiguous signal of the canonical WIMP in the GeV-TeV mass range. The sub-GeV mass range is relatively unexplored due to the difficulty of detecting such light dark matter with traditional techniques. However, there have been recent experimental developments that finally make sub-GeV direct detection viable. I will discuss some of the theoretical principles and strategies to explore sub-GeV dark matter candidates, as well as some current and proposed experimental techniques. I will focus predominantly on semiconductor targets, such as the new SENSEI experiment which utilizes silicon CCDs, and demonstrate the potential for exploring the eV-GeV dark matter mass range in the near future.
"Life after Death: Transient Emission from Compact Objects in Galactic Nuclei"
In most regions of the Universe, stellar orbits have enormous mean free paths, and the timescale for a strong two-body encounter exceeds a Hubble time. However, in dense stellar systems, such as open, globular, and nuclear star clusters, close encounters between stars and/or compact objects are frequent, and may lead to the production of transient electromagnetic or gravitational-wave radiation. I will present my research showing how the densest stellar systems in the Universe — galactic nuclei — are dynamical factories that manufacture transient sources such as X-ray binaries, tidal disruption events, and LIGO-band black hole mergers. I will discuss my past and ongoing work to understand the transient electromagnetic and gravitational radiation from these dynamically assembled systems, focusing especially on ways in which time domain astronomy can probe general relativity.
About the speaker
Nicholas Stone is a NASA Einstein Fellow (2015-2018) at Columbia University, where he works on a variety of topics in theoretical astrophysics. Nicholas' primary collaborators at Columbia are Professors Brian Metzger, Jerry Ostriker, and Zoltan Haiman. Prior to this Nicholas worked as a postdoc at Columbia for two years. He received his PhD in Astronomy and Astrophysics from the Harvard Astronomy department in May 2013. At Harvard, Nicholas was advised by Professor Avi Loeb. Together they worked mainly on the tidal disruption of stars by supermassive black holes. In 2008, Nicholas graduated from from Cornell University with a triple major in Physics, Mathematics, and Economics. From 2000-2004, Nicholas was a student at Montgomery Blair High School.
Brian Humensky's primary research activity is in the area of gamma-ray astrophysics. Brian works on VERITAS, the Very Energetic Radiation Imaging Telescope Array System, and CTA, the Cherenkov Telescope Array (a next-generation instrument). Locally, I collaborate closely with fellow VERITAS and CTA member Reshmi Mukherjee.