Scientific Series

Machine learning has led to recent advancements in image processing, language translation, finance, robotics, musical and visual arts, and medical diagnosis. In this session, we will explore how machine learning can be applied within fields of physics. We will introduce fundamental concepts in machine learning such a neural networks and supervised vs. unsupervised learning, and then proceed to learn to use tools from Python's TensorFlow library.

Bring your laptop. You can attend remotely via Zoom <https://zoom.us/j/154009181>.

Ghostly neutrino particles continue to bring surprises to fundamental physics, from their existence to the phenomenon of neutrino oscillation which implies that their masses are nonzero. Their exact masses, among the most curious unknowns beyond the Standard Model of particle physics, can soon be probed by the joint analysis of upcoming cosmological surveys including LSST, Euclid, WFIRST, Simons Observatory, and CMB-S4. In this talk, I will first discuss ongoing work studying the effects of massive neutrinos. I will then turn the focus to my major efforts of modeling the challenging nonlinear regime of cosmic structures (<10 Mpc) where neutrino effects are the strongest. Finally, I will draw a roadmap to pin down the neutrino mass over the next decade.

We give a quantum speedup for solving the canonical semidefinite programming relaxation for binary quadratic optimization. The class of relaxations for combinatorial optimization has so far eluded quantum speedups. Our methods combine ideas from quantum Gibbs sampling and matrix exponent updates. A de-quantization of the algorithm also leads to a faster classical solver. For generic instances, our quantum solver gives a nearly quadratic speedup over state-of-the-art algorithms.
This is joint work with Fernando Brandao (Caltech) and Daniel Stilck Franca (QMATH, Copenhagen).

The asymptotic symmetry charge algebra of certain class of spacetimes could have a nontrivial central extension, which measures the non-equivariance of the charges of the large gauge transformations. The Cardy formula of the dual CFT has been famously used to derive black hole entropy. However, it remains obscure from the point of view of gravity why such a constant on the gravitational phase space could encode the information about the density of black hole micro-states, and what the degrees of freedom accounting for the black hole entropy truly are. I will discuss the ongoing efforts of understanding these questions in the covariant phase space formalism.

There is a close relationship between derived loop spaces, a geometric object, and Hochschild homology, a categorical invariant, made possible by derived algebraic geometry, thus allowing for both intuitive insights and new computational tools.  In the case of a quotient stack, we discuss a "Jordan decomposition" of loops which is made precise by an equivariant localization result.  We also discuss an Atiyah-Segal completion theorem which relates completed periodic cyclic homology to Betti cohomology.

There is a profound lack of diversity in science labs and classrooms, which has a negative impact on productivity. Scientific research demonstrates that diverse groups are more creative and better able to solve problems. Though the perception is that things are improving, NSERC’s recently released report shows that attrition rates in Canadian STEM fields are higher for women than for men at all career stages and that the percentage of women has not changed substantially in the last 15 years. Racialized and Indigenous people are also underrepresented at Canadian universities. This seminar will explore the scientific research surrounding EDI in hiring, writing reference letters, reviewing grants, and managing trainees. Also included are tips and tricks for incorporating EDI into grant applications.

Many free-fermion topological phases can be diagnosed by analyzing a suitable collection of symmetry data. While the Fu-Kane parity criterion for topological insulators is an early example, the systematic generalization to cover all possible crystalline symmetries and their associated topological phases has only recently been achieved. Interestingly, a key step in the generalization is to recognize the importance of understanding the entanglement-free product states: once these trivial phases are completely mapped, the topological ones are naturally exposed as the “remainders.” In this talk, we will first introduce the theory of symmetry indicators for topological (crystalline) insulators and semimetals, and then discuss how it can be generalized to cover superconductors.

The AdS/CFT correspondence provides a remarkably useful tool
for asking questions in quantum gravity, as it formulates a theory of
quantum gravity in terms of an ordinary non-gravitational quantum field
theory.  Fruitfully exploiting this correspondence therefore requires
understanding how to translate the language of CFT into gravity; a key
insight that has emerged over the past decade is that the entanglement
structure of the CFT must be intimately tied to the emergence of the
gravitational dual (when a classical geometric picture even exists).  In
my talk, I'll review what is currently known about this emergence
(so-called bulk reconstruction), and then discuss some recent results on
precisely how the entanglement of the CFT fixes the geometry on the
gravitational side.  I'll conclude with some open directions, notably
how the result I'll present is amenable to incorporating quantum
corrections.

Making use of modern multi-core computers (and multi-machine supercomputers) in Python can be achieved in a few different ways.  I will present the built-in Python “multiprocessing” module approach, which tends to work well and relatively painlessly for “naturally parallel” (aka “embarrassingly parallel”) problems. I’ll show what is required to use this approach, including how to submit jobs to the Slurm batch system on Symmetry.

Biological evolution is a complex blend of ever changing structural stability, variability and emergence of new phenotypes, niches, ecosystems. We wish to argue that the evolution of life marks the end of a physics world view of law entailed dynamics.