## EE Student Information

### The Department of Electrical Engineering supports Black Lives Matter. Read more.

• • • • •

EE Student Information, Spring Quarter through Academic Year 2020-2021: FAQs and Updated EE Course List.

Updates will be posted on this page, as well as emailed to the EE student mail list.

As always, use your best judgement and consider your own and others' well-being at all times.

# Seminar / Colloquium

## SmartGrid Seminar: Modal Analysis of Power System Data

Topic:
Modal Analysis of Power System Data
Abstract / Description:

Reliable operation of the power grid relies on the real-time stabilization of an interconnected continentalscale network of dynamic components. These include everything from central-station thermal and hydro power plants to end-use loads. The natural uncontrolled responses to disturbances are oscillatory, and swings in voltage and power flows are evident in data from disturbances. Specific control systems are used to damp these natural oscillations that otherwise can result in large-scale blackouts. In order to design and tune controllers it is imperative to understand and anticipate the oscillations, and to develop high fidelity models that can represent this behavior.

In this seminar we focus on the modal analysis of power system data to study oscillations for three purposes:

1. Better the understanding of the oscillations actually present in the grid. (Off-line engineering analysis.)
2. Validation of models used to study the grid. (Operational Planning.)
3. Improve situational awareness by tracking system modes in real-time. (Operations.)

We will provide examples of oscillations observed in the grid in the context of these three objectives. We review the industry standard approaches for power system modal analysis and highlight their shortcomings in practice. They often require preprocessing the data, fine-tuning in model order, and give inconsistent results with related data sets. We suggest that the main attraction of these established methods is that they are computationally easy to use; the calculations require linear algebra. Based on our experience and dissatisfaction with these methods, we explore the use of nonlinear fitting techniques to estimate modes from data. With current computing power, computational barriers are low, and we show excellent results using a variable projection method.

Our method is currently in use by some power engineers to characterize observed oscillations in the West. We show results of using modal analysis to aid in power plant model validation. We discuss and present initial results for adapting our method to estimating modal behavior from ambient data in the grid, which is the focus of our on-going research in this area.

Date and Time:
Wednesday, February 25, 2015 - 1:00pm to 2:00pm
Venue:
Y2E2 270

## Circuitry underlying robust perception and memory

Topic:
Circuitry underlying robust perception and memory
Abstract / Description:

The nervous system is a surprisingly noisy place. For example, if one presents the exact same stimulus to an animal many times, and records the activities of their sensory neurons, those neural responses typically show high levels of trial-to-trial variability. At the same time, we have the experience of having robust thoughts and perceptions. How do our brains generate this robustness from systems of individually unreliable components? In my talk, I will discuss three major circuit mechanisms that have been advanced by my research program. First, I will discuss a mechanism through which the nervous system can learn the statistical structure of the stimuli that it typically experiences. This knowledge allows for the nervous system to make educated guesses about the most likely stimulus presented, even when that stimulus is corrupted by noise. The de-noising process works best when the stimuli, and the noise that corrupts them, have different statistical structures. This motivates the second principle that I will discuss, namely that the peripheral nervous system can shape its noise statistics such that the noise only minimally interferes with the transmission of information about the external world. In particular, I will present a circuit mechanism through which the retina appears to perform this noise shaping. Finally, I will confront the fact that neural systems (like those involved in memory) must typically maintain stable representations even while the responses of individual neurons evolve dynamically over time, and are noisy. Using modern data-science tools, I will illustrate a novel mechanism through which the hippocampus appears to solve this problem. In addition to the importance of these circuit mechanisms for basic science, I will highlight in my talk the implications of my work for the creation of useful biomimetic technologies.

Date and Time:
Monday, March 2, 2015 - 10:00am to 11:00am
Venue:
Clark, S360

## EE Distinguished Lecture: Will Clean Energy Take Our Economy to the Cleaners?

Topic:
Will Clean Energy Take Our Economy to the Cleaners?
Abstract / Description:

In 2009 when the Obama-Biden ticket was inaugurated into office, they set out to accomplish the following aspirational goals:

• Implement an economy-wide cap and trade systems to reduce U.S. Greenhouse Gas Emissions by 83% of 2005 levels by 2050;
• Save more oil than we import from the Middle East and Venezuela by 2019;
• Ensure 10% of our electricity comes from renewables in 2012 and 25% by 2025;
• Put one million Plug-In Hybrid Vehicles on the road by 2015;
• Create five million jobs by investing \$150 bn over ten years to catalyze private efforts to build a clean-energy future.

And in the 2011 State of the Union address, the President put forth a goal to reach 80% clean electricity by 2035. The Office of the Under Secretary for Energy at the US Department of Energy put together a plan to realize these goals, along with the cost of attaining 80% clean electricity by 2035. In this talk, I will summarize this plan, called the "Strategic Technology Energy Plan (STEP)" as well highlight the unique role of run-of-river and pumped storage hydropower in STEP.

Date and Time:
Tuesday, March 10, 2015 - 4:15pm to 5:30pm
Venue:
Hewlett 201

## Applied Physics/Physics Colloquium: Exciton-Polariton based Photonic Circuits

Topic:
Exciton-Polariton based Photonic Circuits
Abstract / Description:

Exciton-Polaritons in semiconductor microcavities are hybrid states of light and matter exhibiting a mix of electronic and photonic properties [1], including strong nonlinearity, low dephasing, ultrafast dynamics, sensitivity to electric and magnetic fields and a rich spin dynamics. These properties are ideal for the construction of a new generation of polaritonic information processing devices, such as exciton-polariton based circuits [2].

For such applications, attention is needed to overcome disorder in the system. Here, advances in the patterning of polariton potentials [3] are theoretically expected to allow for topological polaritons [4]. In analogy to photonic topological insulators [5], propagation immune to scattering with disorder is predicted.

In this presentation, I will also show preliminary work in the application of exciton-polaritons in unconventional circuit architectures such as neural networks [6].

Finally, I will consider the enhancement of nonlinear polaritonic effects at the quantum level with a viewpoint toward triggered single-photon emission [7].

Date and Time:
Wednesday, February 25, 2015 - 2:00pm to 3:00pm
Venue:
Spilker 232

## EE380 Computer Systems Colloquium: A Family of Better Random Number Generators

Topic:
A Family of Better Random Number Generators
Abstract / Description:

Algorithmic random number generators are everywhere, used for all kinds of tasks, from simulation to computational creativity.

Yet most people haven't given much thought to the random number generators they use. Is the RNG you're using a good source of randomness? What does it even mean to be a good RNG?

In this talk, we will examine the desirable properties of a random number generator including performance, correctness, uniformity, and unpredictability, as well as sound mathematical grounding.

We will observe how the RNGs in widespread use lack desirable properties (most commonly failing statistical tests for randomness).

Then we will show how a simple twist on a venerable-but-flawed RNG technique can provide all the properties we desire, resulting in the PCG family of RNGs.

Date and Time:
Wednesday, February 18, 2015 - 4:15pm to 5:15pm
Venue:
Gates B03

## Over the Rainbow: The Other World Seen by Animals

Topic:
OSA Special Seminar: Over the Rainbow: The Other World Seen by Animals
Abstract / Description:

Much of what we understand about the world comes from our eyes, which sense the colors from red to violet that are expressed in the rainbow. Yet we know that this patch of colors is just a small island in the vast electromagnetic spectrum, which extends from radio waves to gamma rays. Two unseen regions of great importance to us are those just over and just under the rainbow - the infrared and ultraviolet, respectively. These were discovered about 200 years ago in inspired experiments that anyone can understand, originally conducted by Frederick William Herschel and Johann Wilhelm Ritter. Only recently has it come to be understood that a variety of animals live in a visual world totally unfamiliar to us, particularly in the ultraviolet.

I will discuss this from the perspective of measurement science, and demonstrate other influences of the ultraviolet in technology, astronomy and climate change.

This seminar is sponsored by Stanford OSA.

Date and Time:
Tuesday, February 17, 2015 - 4:15pm to 5:15pm
Venue:
Spilker 232

## Electroactive Polymers for Integrated Medical Microsystems

Topic:
Electroactive Polymers for Integrated Medical Microsystems
Abstract / Description:

In this talk, I will present the emerging field of electroactive polymer (EAP) microsystems for integrated medical applications, highlighting the EAP actuation mechanisms of soft and hard materials for making passive and active devices: a. electrostriction of a recently developed relaxor ferroelectric polymer, b. diffusion of ions in the case of a hydrogel, and c. dielectric elastomer actuation. We will present the "grand plan" where the variability of polymers, in addition to their intrinsic properties is of great advantage to medical applications. Biocompatibility, light weight, low-cost processing, and flexibility together with operational similarity to biological muscle and the ability to exhibit property changes much beyond what is achievable with inorganic, makes EAPs attractive for applications in artificial muscles, robotics, intelligent medical devices and prosthetics, and drug delivery systems. In each of the aforementioned actuation modes, EAP devices become more efficient as the scale is reduced, providing motivation for downscaling.

While EAPs have the potential to improve many aspects of human life, we will mainly address the applications of flexible micro motors, cardiac tissue engineering and cardiovascular occlusion and the processing and integration challenges that we encountered in working on these systems. Recent advances in polymer microfabrication (i.e. imprint lithography, laser micromachining, and 3D printing), together with breakthroughs in materials science, and understanding of EAP behavior at these small scales will serve to overcome the technological barriers to full integration with microsystems and usher in a new paradigm of medical microsystems.

Date and Time:
Monday, February 9, 2015 - 4:00pm to 5:00pm
Venue:
Allen 101X Auditorium

## Information Systems Lab Colloquium: nderstanding Visual Computations in the Primate Retina

Topic:
nderstanding Visual Computations in the Primate Retina
Abstract / Description:

Vision begins with neural computation in the retina, which sends a highly processed version of the visual world along multiple parallel pathways to the brain. Our research is focused on understanding visual computations in the primate retina and on using this information for the design of artificial retinas to treat blindness. I will describe the state of our understanding of visual computations in the retinal circuitry, with an emphasis on open problems for future exploration, and on achieving a synthetic understanding appropriate for diverse applications.

Date and Time:
Thursday, February 5, 2015 - 4:15pm to 5:15pm
Venue:
Packard 101

## IT-Forum: New matrix decompositions for Gaussian communication networks

Topic:
New matrix decompositions for Gaussian communication networks
Abstract / Description:

A central concept in matrix analysis is the decomposition of a matrix into a product of orthogonal (or unitary) matrices and a diagonal/triangular one, e.g., unitary diagonalization of a symmetric matrix, and more generally the singular-value decomposition, and the QR decomposition. Such decompositions are of particular importance for multi-antenna point-to-point physical-layer communications, where the channel gains are represented by a (channel) matrix. Transforming the channel matrix into diagonal/triangular forms, in this case, allows to reduce the coding task to that of coding for scalar (single-antenna) channels. Thus, the modulation and coding tasks are effectively decoupled and the performance is dictated by the diagonal values. In this work we develop new joint matrix decompositions of several matrices using the same unitary matrix on one side (corresponding to a joint transmitter or receiver) to achieve desired properties for the resulting diagonals. An important special case is a transformation leading to equal diagonals for all matrices simultaneously. This, in turn, allows to construct practical schemes for various communications settings, as well as deriving new theoretic bounds for others.

Date and Time:
Friday, February 6, 2015 - 1:00pm to 2:00pm
Venue:
Packard 202

## SystemX Seminar: Wi-Fi Everywhere: Reflections on Wi-Fi’s (R)Evolution

Topic:
Wi-Fi Everywhere: Reflections on Wi-Fi’s (R)Evolution
Abstract / Description:

In under 2 decades, Wi-Fi has transformed from a standards’ body curiosity to an almost essential convenience in our daily lives. With a unique perspective anchored in a successful Wi-Fi semiconductor company (Atheros), Rick will trace some of this arc in his talk. The presentation will offer an overview of Wi-Fi applications, and the attendant advance of the 802.11 standard. Particular emphasis will then be offered to the supporting technology, notably, the challenges of the past and the opportunities that lie ahead. The talk will close with a broader view of Wi-Fi’s role in the emerging communications landscape.

• Dr. Richard Dasher, SystemX Executive Director, 650-725-3621
• or Ms. Miho Nishi, SystemX Administrative Associate, 650-725-3626​

SystemX Seminars: a new forum for university-industry knowledge exchange.

Date and Time:
Thursday, February 5, 2015 - 4:00pm to 5:00pm
Venue:
Packard 202

## Applied Physics/Physics Colloquium presents "Perspectives on the LHC and new Frontiers in the search for Dark Matter"

Topic:
Perspectives on the LHC and new Frontiers in the search for Dark Matter
Abstract / Description:

By many metrics the Large Hadron Collider has been an astonishing success. It has discovered the mechanism for the origin of mass of fundamental particles and has tested the Standard Model of particle physics over a huge range of scales. However, many of the questions which were open at the start of LHC collisions remain unanswered. In this talk will give a personal perspective on where the LHC has been and where it is going. Looking to the future, I will discuss accelerator-based attempts to solve one of the biggest open questions in particle physics: what is the nature of dark matter?

Date and Time:
Tuesday, December 1, 2020 - 4:30pm
Venue:

## Applied Physics/Physics colloquium presents "Frontiers in Optical and CMB Survey Cosmology"

Topic:
Frontiers in Optical and CMB Survey Cosmology
Abstract / Description:

Observations of the cosmic microwave background (CMB) and the galaxy-filled sky provide images of the universe at its various stages that are sensitive to its physics from the earliest moments to recent times. These observations are key to expanding our understanding to the physics of inflation, neutrinos, dark matter, and dark energy—some of the most mysterious terrains in physics today. We will present results from current CMB and optical surveys—the South Pole Telescope & BICEP/Keck Array and the optical Dark Energy Survey, respectively—whose unprecedented sensitivities enable us to stringently test the standard cosmological model and constrain new physics. We will discuss their current limitations due to calibration uncertainty and confounding astrophysical effects. To conclude, we will look forward to the bright future of optical, CMB, and joint cosmological experiments that will be performed by the Vera C. Rubin Observatory's Legacy Survey of Space and Time and the CMB-S4 experiment.

Date and Time:
Tuesday, November 17, 2020 - 4:30pm

## Applied Physics/Physics colloquium present "Harnessing Data Revolution in Quantum Matter"

Topic:
Harnessing Data Revolution in Quantum Matter
Abstract / Description:

Our desire to better understand quantum emergence drove the community's efforts in improving computing power and experimental instrumentation dramatically. However, the resulting increase in volume and complexity of data present new challenges. I will discuss how these challenges can be embraced and turned into opportunities by employing principled machine learning approaches. The rigorous framework for scientific understanding physicists enjoy through our celebrated tradition requires any machine learning essential to interpret any machine learning. I will discuss our recent results using machine learning approaches designed to be interpretable from the outset. Specifically, I will present discovering order parameters and its fluctuations in voluminous X-ray diffraction data and discovering signature correlations in quantum gas microscopy data as concrete examples.

Date and Time:
Tuesday, November 10, 2020 - 4:30pm

## Applied Physics/Physics Colloquium presents "Advances in our understanding of the airborne transmission of SARS-CoV-2"

Topic:
Advances in our understanding of the airborne transmission of SARS-CoV-2
Abstract / Description:

This talk will focus on the latest research on the transmission of SARS-CoV-2.

Date and Time:
Tuesday, October 27, 2020 - 4:30pm
Venue:

## Applied Physics/Physics Colloquium presents "Discovering the Highest Energy Neutrinos"

Topic:
Discovering the Highest Energy Neutrinos
Abstract / Description:

The detection of high energy astrophysical neutrinos is an important step toward understanding the most energetic cosmic accelerators. IceCube, a large optical detector at the South Pole, has observed the first astrophysical neutrinos and identified at least one potential source. However, the best sensitivity at the highest energies comes from detectors that look for coherent radio Cherenkov emission from neutrino interactions. I will give an overview of the state of current experimental efforts, including recent results, and then discuss a suite of new experiments designed to discover neutrinos at the highest energies and push the energy threshold for radio detection down to overlap with the energy range probed by IceCube, thus covering the full astrophysical energy range out to the highest energies, and opening up new phase space for discovery. These include ground-based experiments such as RNO-G and IceCube-Gen2, as well as the balloon-borne experiment PUEO.

Date and Time:
Tuesday, October 20, 2020 - 4:30pm
Venue:
Zoom ID: 946 67 170 862; +passcode

## AP/Physics colloquium presents "Frontiers in Cosmic Magnetism and in Many-Body Physics"

Topic:
Frontiers in Cosmic Magnetism and in Many-Body Physics
Abstract / Description:

Galaxies like our Milky Way host large-scale, weak magnetic fields. The interstellar magnetic field affects a wide range of physics, from cosmic ray propagation to star formation. The magnetic interstellar medium is also a formidable foreground for experimental cosmology, particularly for the quest to find signatures of inflation in the polarized cosmic microwave background. Despite its importance, the Galactic magnetic field and its role in interstellar processes remain poorly understood. Susan Clark will discuss a few of the big questions that drive her research on cosmic magnetism.

Many-body physics is concerned with the emergent properties - those that characterize the collectivity but not the individual constituents - of macroscopic systems with large numbers of strongly interacting particles. Ensembles of many interacting particles can support qualitatively new phenomena, with the same system able to exist in different universal phases of matter with sharply distinct properties. Due to a variety of conceptual and experimentally motivated reasons, the modern theory of quantum many body systems is largely built around the study of low-temperature and near-equilibrium properties of time independent Hamiltonians. However, such systems represent a small subset of the possible quantum mechanical descriptions of a physical system - which allow for more general unitary evolutions interrupted by non-unitary measurements. Vedika Khemani will describe some highlights of an active research program to advance many-body theory beyond the regime of near-equilibrium time-independent Hamiltonians, with a view towards uncovering complex emergent phenomena in new non-equilibrium regimes. These theoretical efforts are synergistic with recent advances in building controllable quantum devices that naturally implement more general time evolutions generated by circuits of unitary gates, starting from initial states that are not "low energy" in any useful sense.

Date and Time:
Tuesday, October 6, 2020 - 4:30pm
Venue:

## AP 483 Seminar presents "Resolving starlight: a quantum perspective"

Topic:
Resolving starlight: a quantum perspective
Abstract / Description:

The wave-particle duality of light introduces two fundamental problems to imaging, namely, the diffraction limit and the photon shot noise. Quantum information theory can tackle them both in one holistic formalism: model the light as a quantum object, consider any quantum measurement, and pick the one that gives the best statistics. While Helstrom pioneered the theory half a century ago and first applied it to incoherent imaging, it was not until recently that the approach offered a genuine surprise on the age-old topic by predicting a new class of superior imaging methods. For the resolution of two sub-Rayleigh sources, the new methods have been shown theoretically and experimentally to outperform direct imaging and approach the true quantum limits. Recent efforts to generalize the theory for an arbitrary number of sources suggest that, despite the existence of harsh quantum limits, the quantum-inspired methods can still offer significant improvements over direct imaging for subdiffraction objects, potentially benefiting many applications in astronomy as well as fluorescence microscopy.

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Date and Time:
Monday, November 16, 2020 - 4:15pm

## AP 483 Seminar presents "Correlated photon physics in mesoscopic atomic chains"

Topic:
Correlated photon physics in mesoscopic atomic chains
Abstract / Description:

Tightly packed ordered arrays of atoms (or, more generally, quantum emitters) exhibit remarkable collective optical properties, as dissipation in the form of photon emission is correlated. In this talk, I will discuss the single-, few- and many-body out-of-equilibrium physics of 1D arrays, and their potential to realize versatile light-matter interfaces. For small enough inter-atomic distances, atomic chains feature dark states that allow for dissipationless transport of photons, behaving as waveguides for single-photon states. Atomic waveguides can be used to mediate interactions between impurity qubits coupled to the array, and allow for the realization of multiple paradigms in waveguide QED, from bandgap physics to chiral quantum optics [1]. Due to the two-level nature of the atoms, atomic waveguides are a perfect playground to realize strong photon-photon interactions. At the many-body level, I will address the open question of how the geometry of the array impacts the process of "Dicke superradiance", where fully inverted atoms synchronize as they de-excite, emitting light in a burst (in contrast to the exponential decay expected from independent emitters). While most literature attributes the quenching of superradiance to Hamiltonian dipole-dipole interactions, the actual culprits are dissipative processes in the form of photon emission into different optical modes. I will provide an understanding of the physics in terms of collective jump operators and demonstrate that superradiance survives at small inter-atomic distances [2]. I will finish my talk by discussing the implications of correlated photon emission for quantum information processing and metrology.

[1] S. J. Masson, A. Asenjo-Garcia, Atomic-waveguide Quantum Electrodynamics, arXiv: 1912.06234 (2019)

[2] S. J. Masson, I. Ferrier-Barbut, L. A. Orozco, A. Browaeys, A. Asenjo-Garcia, Many-body signatures of collective decay in atomic chains, arXiv: 2008.08139 (2020)

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Date and Time:
Monday, November 9, 2020 - 4:15pm

## AP 483 Seminar presents "Benchmarking quantum computers and future directions for superconducting quantum hardware"

Topic:
Benchmarking quantum computers and future directions for superconducting quantum hardware
Abstract / Description:

While the fully fault-tolerant universal quantum computing system is still many years ahead, building an early quantum computer with quantum advantage becomes a feasible near-term milestone that we can realistically plan. Increasing number of near-term applications has been accelerating the development of quantum hardware in the industries, and as quantum system size grows, we need a whole system metric to evaluate the level of hardware performance. I would like to introduce the quantum volume (arXiv:1811.12926 and more recently arXiv:2008.08571) as a system-level metric that quantifies quantum computational power of early quantum computing processors. The quantum volume depends on various individual component metrics such as gate fidelity and crosstalk. I will discuss some of the challenges in building superconducting quantum hardware and suggest few directions to improve the quantum volume.

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Date and Time:
Monday, November 2, 2020 - 4:15pm

## AP 483 Seminar presents "Quantum Technologies Enabled by Cavity-Optomechanics at Low Temperatures"

Topic:
Quantum Technologies Enabled by Cavity-Optomechanics at Low Temperatures
Abstract / Description:

The ability to fabricate devices that demonstrate quantum behavior, as opposed to being restricted to what nature has given us, has opened up the possibility of technologies based on the laws of quantum physics instead of classical physics. The prime example of this is the superconducting qubit at the core of a quantum processor. Another class of fabricated devices that are beginning to demonstrate quantum behavior is that of nanoscale mechanical resonators. Such devices promise force and torque sensors operating at or beyond the standard quantum limit, and novel technologies such as quantum-level wavelength transducers. I will tell you about our efforts to develop such technologies and some of the fun physics we have uncovered along the way.

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Date and Time:
Monday, October 26, 2020 - 4:15pm
Venue:
Zoom

## CS300 Seminar

Topic:
Abstract / Description:

Individuals of all genders invited to be a part of:
#StanfordToo: A Conversation about Sexual Harassment in Our Academic Spaces, where we will feature real stories of harassment at Stanford academic STEM in a conversation with Provost Drell, Dean Minor (SoM), and Dean Graham (SE3). We will have plenty of time for audience discussion on how we can take concrete action to dismantle this culture and actively work towards a more inclusive Stanford for everyone. While our emphasis is on STEM fields, we welcome and encourage participation from students, postdocs, staff, and faculty of all academic disciplines and backgrounds.

Date and Time:
Friday, April 19, 2019 - 3:30pm
Venue:
STLC 111

## John G. Linvill Distinguished Seminar on Electronic Systems Technology

Topic:
Internet of Things and Internet of Energy for Connecting at Any Time and Any Place
Abstract / Description:

In this presentation, I would like to discuss with you how to establish a sustainable and smart society through the internet of energy for connecting at any time and any place. I suspect that you have heard the phrase, "Internet of Energy" less often. The meaning of this phrase is simple. Because of a ubiquitous energy transmission system, you do not need to worry about a shortage of electric power. One of the most important items for establishing a sustainable society is [...]

"Inaugural Linvill Distinguished Seminar on Electronic Systems Technology," EE News, July 2018

Date and Time:
Monday, January 14, 2019 - 4:30pm
Venue:
Hewlett 200

## Special Seminar: Formal Methods meets Machine Learning: Explorations in Cyber-Physical Systems Design

Topic:
Formal Methods meets Machine Learning: Explorations in Cyber-Physical Systems Design
Abstract / Description:

Cyber-physical systems (CPS) are computational systems tightly integrated with physical processes. Examples include modern automobiles, fly-by-wire aircraft, software-controlled medical devices, robots, and many more. In recent times, these systems have exploded in complexity due to the growing amount of software and networking integrated into physical environments via real-time control loops, as well as the growing use of machine learning and artificial intelligence (AI) techniques. At the same time, these systems must be designed with strong verifiable guarantees.

In this talk, I will describe our research explorations at the intersection of machine learning and formal methods that address some of the challenges in CPS design. First, I will describe how machine learning techniques can be blended with formal methods to address challenges in specification, design, and verification of industrial CPS. In particular, I will discuss the use of formal inductive synthesis --- algorithmic synthesis from examples with formal guarantees — for CPS design. Next, I will discuss how formal methods can be used to improve the level of assurance in systems that rely heavily on machine learning, such as autonomous vehicles using deep learning for perception. Both theory and industrial case studies will be discussed, with a special focus on the automotive domain. I will conclude with a brief discussion of the major remaining challenges posed by the use of machine learning and AI in CPS.

Date and Time:
Monday, December 4, 2017 - 4:00pm
Venue:
Gates 463A

## SpaceX's journey on the road to mars

Topic:
SpaceX's journey on the road to mars
Abstract / Description:

SSI will be hosting Gwynne Shotwell — President and COO of SpaceX — to discuss SpaceX's journey on the road to mars. The event will be on Wednesday Oct 11th from 7pm - 8pm in Dinkelspiel Auditorium. After the talk, there will be a Q&A session hosted by Steve Jurvetson from DFJ Venture Capital.

Claim your tickets now on eventbright

Date and Time:
Wednesday, October 11, 2017 - 7:00pm
Venue:
Dinkelspiel Auditorium

## CS Department Lecture Series (CS300)

Topic:
Faculty speak about their research to new PhD students
Abstract / Description:

Offered to incoming first-year PhD students in the Autumn quarter.

The seminar gives CS faculty the opportunity to speak about their research, which allows new CS PhD students the chance to learn about the professors and their research before permanently aligning.

4:30-5:15, Subhasish Mitra

5:15-6:00, Silvio Savarese

Date and Time:
Wednesday, December 7, 2016 - 4:30pm to 6:00pm
Venue:
200-305 Lane History Corner, Main Quad

## CS Department Lecture Series (CS300)

Topic:
Faculty speak about their research to new PhD students
Abstract / Description:

Offered to incoming first-year PhD students in the Autumn quarter.

The seminar gives CS faculty the opportunity to speak about their research, which allows new CS PhD students the chance to learn about the professors and their research before permanently aligning.

4:30-5:15, Phil Levis

5:15-6:00, Ron Fedkiw

Date and Time:
Monday, December 5, 2016 - 4:30pm to 6:00pm
Venue:
200-305 Lane History Corner, Main Quad

## CS Department Lecture Series (CS300)

Topic:
Faculty speak about their research to new PhD students
Abstract / Description:

Offered to incoming first-year PhD students in the Autumn quarter.

The seminar gives CS faculty the opportunity to speak about their research, which allows new CS PhD students the chance to learn about the professors and their research before permanently aligning.

4:30-5:15, Dan Boneh

5:15-6:00, Aaron Sidford

Date and Time:
Wednesday, November 30, 2016 - 4:30pm to 6:00pm
Venue:
200-305 Lane History Corner, Main Quad

## CS Department Lecture Series (CS300)

Topic:
Faculty speak about their research to new PhD students
Abstract / Description:

Offered to incoming first-year PhD students in the Autumn quarter.

The seminar gives CS faculty the opportunity to speak about their research, which allows new CS PhD students the chance to learn about the professors and their research before permanently aligning.

4:30-5:15, John Mitchell

5:15-6:00, James Zou

Date and Time:
Monday, November 28, 2016 - 4:30pm to 6:00pm
Venue:
200-305 Lane History Corner, Main Quad

## CS Department Lecture Series (CS300)

Topic:
Faculty speak about their research to new PhD students
Abstract / Description:

Offered to incoming first-year PhD students in the Autumn quarter.

The seminar gives CS faculty the opportunity to speak about their research, which allows new CS PhD students the chance to learn about the professors and their research before permanently aligning.

4:30-5:15, Emma Brunskill

5:15-6:00, Doug James

Date and Time:
Wednesday, November 16, 2016 - 4:30pm to 6:00pm
Venue:
200-305 Lane History Corner, Main Quad

## CS Department Lecture Series (CS300)

Topic:
Faculty speak about their research to new PhD students
Abstract / Description:

Offered to incoming first-year PhD students in the Autumn quarter.

The seminar gives CS faculty the opportunity to speak about their research, which allows new CS PhD students the chance to learn about the professors and their research before permanently aligning.

4:30-5:15, James Landay

5:15-6:00, Dan Jurafsky

Date and Time:
Monday, November 14, 2016 - 4:30pm to 6:00pm
Venue:
200-305 Lane History Corner, Main Quad

## SystemX presents "Quantum Computing – Where does the Hope meet the Hype?"

Topic:
Quantum Computing – Where does the Hope meet the Hype?
Abstract / Description:

Quantum computers exploit the bizarre features of quantum physics -- uncertainty, entanglement, and measurement -- to perform tasks that are impossible using conventional means, such as computing over ungodly amounts of data, and communicating via teleportation. I will describe the architecture of a quantum computer based on individual atomic clock qubits, suspended and isolated with electric fields, perfectly replicable with no idle errors, and individually addressed with laser beams. This leading physical representation of a quantum computer has allowed unmatched demonstrations of small algorithms and emulations of hard quantum problems with more than 50 quantum bits. While this system can solve some esoteric tasks that cannot be accomplished in conventional devices, it remains a great engineering challenge to build a quantum computer big enough to be generally useful for society. But the good news is that this is not a scientific challenge, as we know the technology needed and it's not quantum.

Date and Time:
Thursday, October 8, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## EE380 Computer Systems Colloquium presents "Coming Attractions: Death or Utopia in the Next Three Decades"

Topic:
Coming Attractions: Death or Utopia in the Next Three Decades
Abstract / Description:

Today the data suggests that we are near the beginning of a chaotic mess of global proportions. Things are fairly simple: a global pandemic with no tools to fight the virus. a global economy in disarray, climate change and other existential risks beginning to intrude into our daily lives, and a total lack of a plan as to what to do. On the other hand, we are at the pinnacle of human capabilities and have, if we choose to do so, the capability create a Utopian egalitarian world without conflict or want.

In this two hour program a group of experts will explore the future, focusing on 2030 and 2050. Where are we now? What is trending? What if anything can be done about it?

Sponsors: The invitational Asilomar Microcomputer Workshop is one of the iconic gatherings which supported the growth of computing. This is the first mini-conference which replaces the 46th Asilomar Microcomputer Workshop which was canceled due to the COVID-19 pandemic. www.amw.org.

Date and Time:
Thursday, June 4, 2020 - 11:00am
Venue:

## EE380 Computer Systems Colloquium presents "Rebooting the Internet"

Topic:
Rebooting the Internet
Abstract / Description:

"Build one and throw it out," so the adage goes. This talk explores the argument for systems architecture revision, including processors, operating systems, and networking--both as a general principle and the ways the Internet in particular is currently in need of a reboot. Assumptions, resources, and goals change with time and experience, and so too does our understanding of architectural principles. Through the eyes of the Internet and other examples, we review what we got right (one ring to rule them all, good enough rather than perfect), what we got wrong (7 layers, name resolution as afterthought), and what we only now are beginning to appreciate (layering and forwarding as one thing). Challenge cases are presented that can help drive this redesign, including single-packet exchanges and recursive layering, and an example given of one direction this approach can lead. Finally, we explore the challenges of evolution and transition to help us prepare for giving the Internet a well-deserved reset.

REMOTE URL SU-EE380-20200311

Date and Time:
Wednesday, March 11, 2020 - 4:30pm
Venue:
[remote only]

## EE380 Computer Systems Colloquium presents "Data Analytics at the Exascale for Free Electron Lasers Lasers Project"

Topic:
Abstract / Description:

The increase in velocity, volume, and complexity of the data generated by the upcoming Linac Coherent Light Source upgrade (LCLS-II) at the SLAC National Accelerator Laboratory presents a considerable challenge for data acquisition, data processing, and data management. These systems face formidable challenges due to the extremely high data throughput, hundreds of GB/s to multi-TB/s, generated by the detectors at the experimental facilities and to the intensive computational demand for data processing and scientific interpretation. The LCLS-II Data System is a fast, powerful, and flexible architecture that includes a feature extraction layer designed to reduce the data volumes by at least one order of magnitude while preserving the science content of the data. Innovative architectures are required to implement this reduction with a configurable approach that can adapt to the multiple science areas served by LCLS. In order to increase the likelihood of experiment success and improve the quality of recorded data, a real-time analysis framework provides visualization and graphically-configurable analysis of a selectable subset of the data on the timescale of seconds. A fast feedback layer offers dedicated processing resources to the running experiment in order to provide experimenters feedback about the quality of acquired data within minutes. We will present an overview of the LCLS-II Data System architecture with an emphasis on the Data Reduction Pipeline (DRP) and online monitoring framework.

Date and Time:
Wednesday, March 4, 2020 - 4:30pm
Venue:
Shriram 104

## EE380 Computer Systems Colloquium presents "The Soul of a New Machine: Rethinking the Computer"

Topic:
The Soul of a New Machine: Rethinking the Computer
Abstract / Description:

While our software systems have become increasingly elastic, the physical substrate available to run that software (that is, the computer!) has remained stuck in a bygone era of PC architecture. Hyperscale infrastructure providers have long since figured this out, building machines that are fit to purpose -- but those advances have been denied to the mass market. In this talk, we will talk about our vision for a new, rack-scale, server-side machine -- and how we anticipate advances like open firmware, RISC-V, and Rust will play a central role in realizing that vision.

Date and Time:
Wednesday, February 26, 2020 - 4:30pm
Venue:
Shriram 104

## EE380 Computer Systems Colloquium presents "fastai: A Layered API for Deep Learning"

Topic:
fastai: A Layered API for Deep Learning
Abstract / Description:

[speaker photo] Sylvain's research at fast.ai has focused on designing and improving techniques that allow models to train fast with limited resources. He has also been a core developer of the fastai library, including implementing the warping transformations, the preprocessing pipeline, much of fastai.text, and a lot more.
Prior to fastai, Sylvain was a Mathematics and Computer Science teacher in Paris for seven years. He taught CPGE, the 2-year French program that prepares students for graduate programs at France's top engineering schools (the "grandes écoles"). After relocating to the USA in 2015, Sylvain wrote ten textbooks covering the entire CPGE curriculum. Sylvain is an alumni from École Normale Supérieure (Paris, France) and has a Master's Degree in Mathematics from University Paris XI (Orsay, France). He lives in Brooklyn with his husband and two sons.fastai is a deep learning library which provides practitioners with high-level components that can quickly and easily provide state-of-the-art results in standard deep learning domains, and provides researchers with low-level components that can be mixed and matched to build new approaches. It aims to do both things without substantial compromises in ease of use, flexibility, or performance. This is possible thanks to a carefully layered architecture, which expresses common underlying patterns of many deep learning and data processing techniques in terms of decoupled abstractions. These abstractions can be expressed concisely and clearly by leveraging the dynamism of the underlying Python language and the flexibility of the PyTorch library. fastai includes: a new type dispatch system for Python along with a semantic type hierarchy for tensors; a GPU-optimized computer vision library which can be extended in pure Python; an optimizer which refactors out the common functionality of modern optimizers into two basic pieces, allowing optimization algorithms to be implemented in 4-5 lines of code; a novel 2-way callback system that can access any part of the data, model, or optimizer and change it at any point during training; a new data block API; and much more. We have used this library to successfully create a complete deep learning course, which we were able to write more quickly than using previous approaches, and the code was more clear. The library is already in wide use in research, industry, and teaching.

Date and Time:
Wednesday, February 19, 2020 - 4:30pm
Venue:
Shriram 104

## EE380 Computer Systems Colloquium presents "Centaur Technology's Deep learning Coprocessor"

Topic:
Centaur Technology's Deep learning Coprocessor
Abstract / Description:

This talk will explore Centaur's microprocessor-design technology includes both a new high-performance x86 core AND the industry's FIRST integrated AI Coprocessor for x86 systems. The x86 microprocessor cores deliver high instructions/clock (IPC) for server-class applications and support the latest x86 extensions such as AVX 512 and new instructions for fast transfer of AI data. The AI Coprocessor is a clean-sheet processor designed to deliver high performance and efficiency on deep-learning applications, freeing up the x86 cores for general-purpose computing. Both the x86 and AI Coprocessor technologies have now been proven in silicon using a new scalable SoC platform with eight x86 cores and an AI Coprocessor able to compute 20 trillion AI operations/sec with 20 terabytes/sec memory bandwidth. This SoC architecture requires less than 195mm2 in TSMC 16nm and provides an extensible platform with 44 PCIe lanes and 4 channels of PC3200 DDR4.

Date and Time:
Wednesday, February 12, 2020 - 4:30pm
Venue:
Shriram 104

## EE380 Computer Systems Colloquium presents "KUtrace 2020"

Topic:
KUtrace 2020
Abstract / Description:

Observation tools for understanding occasionally-slow performance in large-scale distributed transaction systems are not keeping up with the complexity of the environment. The same applies to large database systems, to real-time control systems in cars and airplanes, and to operating system design.

Extremely low-overhead tracing can reveal the true execution and non-execution (waiting) dynamics of such software, running in situ with live traffic. KUtrace is such a tool, based on small Linux kernel patches recording and timestamping every transition between kernel- and user-mode execution across all CPUs of a datacenter or vehicle computer. The resulting displays show exactly what each transaction is doing every nanosecond, and hence shows why unpredictable ones are slow, all with tracing overhead well under 1%. Recent additions to KUtrace also show interference between programs and show profiles within long execution stretches that have no transitions.

The net result is deep insight into the dynamics of complex software, leading to often-simple changes to improve performance.

Date and Time:
Wednesday, February 5, 2020 - 4:30pm
Venue:
Shriram 104

## EE380 Computer Systems Colloquium presents "Computer Security: The Mess We're In, How We Got Here, and What to Do About It"

Topic:
Computer Security: The Mess We're In, How We Got Here, and What to Do About It
Abstract / Description:

There are dozens of companies selling products to solve your Identity and Access Management (IAM) problems. Why so many? Because IAM is really hard. But why is IAM so hard? I hope to convince you in this talk the reason is that our approach to IAM is fundamentally flawed. The talk includes my personal theory of how we got into this mess and a strategy for getting out of it.

Date and Time:
Wednesday, January 29, 2020 - 4:30pm
Venue:
Shriram 104

## EE380 Computer Systems Colloquium presents "A Fight over the Law of Software APIs, and other war stories from the Electronic Frontier Foundation""

Topic:
A Fight over the Law of Software APIs, and other war stories from the Electronic Frontier Foundation
Abstract / Description:

Over the past four years, a federal appeals court has upended a basic assumption about software: that while software can be covered by copyright law, software interfaces cannot. The court ruled that Google's re-implementation of Java APIs infringed Oracle's Java copyrights. Now, the Supreme Court will weigh in. The Electronic Frontier Foundation has argued at each stage of the case that copyright in APIs threatens innovation, especially by smaller players. This is part of EFF's work promoting the freedom to tinker, innovate, and research, especially outside the walls of major corporations—work that has also included challenges to overbroad computer crime and "anti-circumvention" laws.

Date and Time:
Wednesday, January 22, 2020 - 4:30pm
Venue:
Shriram 104

## OSA/SPIE, SPRC and Ginzton Lab present "Next Generation Photonics"

Topic:
Next Generation Photonics
Abstract / Description:

Over the past few decades, silicon photonics has revolutionized photonic integrated circuits by leveraging the semiconductor CMOS manufacturing infrastructure for low cost, high performance devices and systems. However, key fundamental challenges of the field remain unsolved: packaging the devices with optical fibers and generating light on chip.

We developed a novel approach for fiber packaging based on fusing the fiber and chip together. Connecting a silicon photonic chip across long distances requires attaching optical fibers to the chip. In practice, the packaging of optical fibers to photonic devices is time consuming, lossy, and expensive. This process is usually done by gluing the fiber and chip together using optical adhesives. By fusion splicing the chip and fiber together we have demonstrated losses as low as 1dB for single fibers and 2.5dB for an array of four fibers.

Imagine a laser as thick as an atom that is compatible with silicon photonics. Silicon based materials are passive. Since silicon is an indirect band gap material it is a poor light emitter. To generate light on a silicon photonic integrated circuit, you need to integrate active materials, which are usually not compatible with CMOS, with the device. Two-dimensional materials are excellent candidates for light sources, modulators, and detectors; and they are compatible with CMOS electronic manufacturing in the back end. We recently demonstrated the first fully on-chip 2D laser. Due to their thickness and transfer process, we envision electronic-photonic devices with many optical layers, each with their own lasers, modulators, and detectors based on 2D materials. Our recent demonstration completes the set of active devices completely based on 2D materials.

Date and Time:
Wednesday, May 6, 2020 - 1:00pm
Venue:
Zoom ID: 991 2584 9377

## OSA/SPIE with SPRC and Ginzton Lab present "Next Generation Photonics"

Topic:
Next Generation Photonics
Abstract / Description:

Over the past few decades, silicon photonics has revolutionized photonic integrated circuits by leveraging the semiconductor CMOS manufacturing infrastructure for low cost, high performance devices and systems. However, key fundamental challenges of the field remain unsolved: packaging the devices with optical fibers and generating light on chip.

We developed a novel approach for fiber packaging based on fusing the fiber and chip together. Connecting a silicon photonic chip across long distances requires attaching optical fibers to the chip. In practice, the packaging of optical fibers to photonic devices is time consuming, lossy, and expensive. This process is usually done by gluing the fiber and chip together using optical adhesives. By fusion splicing the chip and fiber together we have demonstrated losses as low as 1dB for single fibers and 2.5dB for an array of four fibers.

Imagine a laser as thick as an atom that is compatible with silicon photonics. Silicon based materials are passive. Since silicon is an indirect band gap material it is a poor light emitter. To generate light on a silicon photonic integrated circuit, you need to integrate active materials, which are usually not compatible with CMOS, with the device. Two-dimensional materials are excellent candidates for light sources, modulators, and detectors; and they are compatible with CMOS electronic manufacturing in the back end. We recently demonstrated the first fully on-chip 2D laser. Due to their thickness and transfer process, we envision electronic-photonic devices with many optical layers, each with their own lasers, modulators, and detectors based on 2D materials. Our recent demonstration completes the set of active devices completely based on 2D materials.

Date and Time:
Monday, May 4, 2020 - 12:05pm
Venue:
Zoom ID: 991 2584 9377 (contact organizers for password)

## OSA/SPIE, SPRC and Ginzton Lab present "Applications of Precision Motion Control in the Optics and Photonics Industry"

Topic:
Applications of Precision Motion Control in the Optics and Photonics Industry
Abstract / Description:

Precision Automation has become a rapidly growing industry as manufacturers must develop ways to achieve ever-tightening tolerances. Likewise, research must push the envelope in order to maintain the increasingly advancing pace of technology. Aerotech has been providing assistance and solutions to researchers within academia and laboratories worldwide for 50 years and continues to empower students and technicians to complete and publish their research. In this seminar, we will discuss why automation and motion control is important as well as the various methods by which Aerotech has helped expand the boundaries of research in the field of optics and photonics. The topics we will cover in this session include:

• What Precision Motion Control is and why it is important
• How to choose the right set of motion control
• How Aerotech has helped further research in:
• Optics design and manufacturing
• Light sources and beamlines
• Optical positioning and ultra-fast spectroscopy
• Fiber alignment
• Laser Processing

After the presentation, a round-table discussion will be held during which we'd like to hear from you! Please bring with you any topics and - more importantly - any tough issues that you'd like to pose pertaining to automated motion control within your specific research field.

Date and Time:
Tuesday, March 3, 2020 - 4:15pm
Venue:
Spilker 232

## OSA/SPIE present "Applications of Precision Motion Control in the Optics and Photonics Industry"

Topic:
Applications of Precision Motion Control in the Optics and Photonics Industry
Abstract / Description:

Refreshments at 4:00

Precision Automation has become a rapidly growing industry as manufacturers must develop ways to achieve ever-tightening tolerances. Likewise, research must push the envelope in order to maintain the increasingly advancing pace of technology. Aerotech has been providing assistance and solutions to researchers within academia and laboratories worldwide for 50 years and continues to empower students and technicians to complete and publish their research. In this seminar, we will discuss why automation and motion control is important as well as the various methods by which Aerotech has helped expand the boundaries of research in the field of optics and photonics. The topics we will cover in this session include:

• What Precision Motion Control is and why it is important
• How to choose the right set of motion control
• How Aerotech has helped further research in:
• Optics design and manufacturing
• Light sources and beamlines
• Optical positioning and ultra-fast spectroscopy
• Fiber alignment
• Laser Processing

After the presentation, a round-table discussion will be held during which we'd like to hear from you! Please bring with you any topics and - more importantly - any tough issues that you'd like to pose pertaining to automated motion control within your specific research field.

Date and Time:
Wednesday, February 26, 2020 - 4:15pm
Venue:
Spilker 232

## OSA/SPIE, SPRC and Ginzton Lab present "Innovation in the Laser Industry"

Topic:
Innovation in the Laser Industry
Abstract / Description:
Spectra-Physics was a startup when Prof. Byer was a summer student in 1964.  By the time Alan Petersen and Jim Kafka joined in the early 1980s, the company had matured and aspired to be all things in lasers.  We will discuss the changes in the laser industry since that time and the role of innovation within the company to drive growth and broaden the product offerings from scientific lasers to industrial laser systems.

PLEASE RVP TO GUARANTEE A SPOT.
Date and Time:
Monday, March 9, 2020 - 12:30pm
Venue:
Y2E2 299

## Q-Farm Quantum Seminar Series presents "Surprises from Time Crystals"

Topic:
Surprises from Time Crystals
Abstract / Description:

Time crystals are new states of matter that only exist in an out-of-equilibrium setting. I will review the state of this rapidly evolving field, focusing in particular on some of the remarkable properties of this phase, and the surprises coming out of its study. I will provide a detailed overview of existing experiments, with a view towards identifying the ingredients needed for an unambiguous observation of this phase in the future.

Date and Time:
Wednesday, January 29, 2020 - 12:00pm
Venue:
Hansen Physics & Astrophysics Building, 102/103

## AP 483 Seminar Series presents "Photovoltaic Restoration of Sight in Retinal Degeneration"

Topic:
Photovoltaic Restoration of Sight in Retinal Degeneration
Abstract / Description:

Retinal degenerative diseases lead to blindness due to loss of the “image capturing” photoreceptors, while neurons in the “image-processing” inner retinal layers are relatively well-preserved. Information can be reintroduced into the visual system by photovoltaic subretinal implants, which convert incident light into electric current and stimulate the secondary retinal neurons.

To provide sufficient light intensity for photovoltaic stimulation while avoiding visual perception by remaining photoreceptors, images captured by a camera are projected onto the retina from augmented-reality glasses using pulsed near-infrared light. This design avoids the use of bulky electronics and wiring, thereby greatly reducing the surgical complexity and enabling scaling the number of photovoltaic pixels to thousands. Many features of the natural retinal signal processing are preserved in this approach, and spatial resolution matches the pixel pitch (so far 100 μm pixels in human patients, and 50 μm in rodents). For a broad acceptance of this technology by patients who lost central vision due to Age-Related Macular Degeneration, visual acuity should exceed 20/100, which requires pixels smaller than 25 μm. I will present a 3-dimensional electro-neural interface scalable to cellular dimensions and discuss the outlook and challenges for future developments.

AP483 Optics and Electronics Seminar Series 2019-20 (Sponsored by Ginzton Laboratory, SPRC, Applied Physics, Physics, and HEPL).

Date and Time:
Monday, February 24, 2020 - 4:15pm
Venue:
Spilker 232

## AP 483 Seminar Series presents "Taking the Humble FBG on a Voyage of Discovery from the Lab Bench to the Hospital and Beyond"

Topic:
Taking the Humble FBG on a Voyage of Discovery from the Lab Bench to the Hospital and Beyond
Abstract / Description:

Fiber Bragg gratings (FBGs) emerged almost magically in 1978. Since then, they have developed from being primarily of academic interest, to being one of the most versatile photonic components for both telecommunications and sensing.

I was lucky enough to get introduced to photonics during the early years of discovery and experimentation with FBGs and ended up making FBG components for telecommunications. Jumping forward a few decades, I had moved away from telecommunications and had started to dabble with free space optical sensing. Then, during a chance meeting over a beer in Sydney, I was asked, rather naively, if it was possible to use optical fiber to monitor what goes on in the esophagus when we swallow. This set me off on an entirely new path and, pulling together some ideas from telecoms, and some basic mechanical engineering, my team ended up developing a range of sensors for monitoring pressure in the human digestive tract.

Being able to detect swallowing disorders quickly led to monitoring in other regions of the gut, like the colon and small bowel, and, together with colleagues from Flinders Medical Center, we provided some details of the inner workings of the human gastrointestinal tract. Thus, the fiber optic catheter was born. This device kept me busy for almost 10 years, during which time we worked closely with clinical research groups to write a whole new chapter on how the gut works.

The next ‘Eureka’ moment came when we had to develop a temperature independent version of our sensor, initially for monitoring pressure beneath bandages. The very simple design that resulted worked well for sub-bandage measurements but has also become a key technology that has moved to applications in aerospace, pipeline monitoring, and mining.

When I started working in optics, I never thought I'd end up monitoring what makes your stomach growl when hungry, how air flows across an airplane wing, or detecting pressure transients in water pipes.

During this talk I will explain how our basic transducers work and will then describe the applications they are now being applied to, demonstrating how the humble FBG has opened up the scope and reach of fiber-optic sensing.

AP483 Optics and Electronics Seminar Series 2019-20 (Sponsored by Ginzton Laboratory, SPRC, Applied Physics, Physics, and HEPL).

Date and Time:
Monday, February 10, 2020 - 4:15pm
Venue:
Spilker 232

## AP 483 Seminar Series presents "Non-Hermitian Photonics: Optics at an Exceptional Point"

Topic:
Non-Hermitian Photonics: Optics at an Exceptional Point
Abstract / Description:

In recent years, non-Hermitian degeneracies, also known as exceptional points (EPs), have emerged as a new paradigm for engineering the response of optical systems. At such points, an N-dimensional space can be represented by a single eigenvalue and one eigenvector. As a result, these points are associated with abrupt phase transitions in parameter space. Among many different non-conservative photonic configurations, parity-time (PT) symmetric systems are of particular interest since they provide a powerful platform to explore, and consequently utilize, the physics of exceptional points in a systematic manner. In this talk, I will review some of our recent works in the area of non-Hermitian (mainly PT-symmetric) active photonics. For example, in a series of works, we have demonstrated how the generation and judicial utilization of these points in laser systems can result in unexpected dynamics, unusual linewidth behavior, and improved modal response. On the other hand, biasing a photonic system at an exceptional point can lead to orders of magnitude enhancement in sensitivity, an effect that may enable a new generation of ultrasensitive optical sensors on-chip. Non-Hermiticity can also be used as a means to promote or single out an edge mode in photonic topological insulator lattices. Rotation sensors play a crucial role in a diverse set of applications associated with navigation, positioning, and inertial sensing. Most optical gyroscopes rely on the Sagnac effect induced phase shift that scales linearly with the rotational velocity. In ring laser gyroscopes (RLGs), this shift manifests itself as a resonance splitting in the emission spectrum that can be detected as a beat frequency. The need for evermore precise RLGs has fueled research activities towards devising new approaches aimed to boost the sensitivity beyond what is dictated by geometrical constraints. In this respect, attempts have been made in the past to use either dispersive or nonlinear effects. Here, we propose a new scheme for ultrasensitive laser gyroscopes that utilizes the physics of exceptional points. By exploiting the properties of such non- Hermitian degeneracies, we show that the rotation-induced frequency splitting becomes proportional to the square root of the gyration speed, thus enhancing the sensitivity to low angular rotations by orders of magnitudes. We will then describe a possible modification of a standard RLG to support an exceptional point and measure the resulting enhanced sensitivity in the proposed system.

Date and Time:
Monday, January 27, 2020 - 4:15pm
Venue:
Spilker 232

## AP 483 Seminar Series presents "Where Are We Heading: A Brief History and Future of Navigation"

Topic:
Abstract / Description:

- tba -

AP483 Optics and Electronics Seminar Series 2019-20 (Sponsored by Ginzton Laboratory, SPRC, Applied Physics, Physics, and HEPL).

Date and Time:
Monday, January 13, 2020 - 4:15pm
Venue:
Spilker 232

## ISL Colloquium presents "Learning Convolutions from Scratch"

Topic:
Learning Convolutions from Scratch
Abstract / Description:

Convolution is one of the most essential components of architectures used in computer vision. As machine learning moves towards reducing the expert bias and learning it from data, a natural next step seems to be learning convolution-like structures from scratch. This, however, has proven elusive. For example, current state-of-the-art architecture search algorithms use convolution as one of the existing modules rather than learning it from data. In an attempt to understand the inductive bias that gives rise to convolutions, we investigate minimum description length as a guiding principle and show that in some settings, it can indeed be indicative of the performance of architectures. To find architectures with small description length, we propose β-LASSO, a simple variant of LASSO algorithm that, when applied on fully-connected networks for image classification tasks, learns architectures with local connections and achieves state-of-the-art accuracies for training fully-connected nets on CIFAR-10 (85.19%), CIFAR-100 (59.56%) and SVHN (94.07%) bridging the gap between fully-connected and convolutional nets.

Date and Time:
Thursday, November 5, 2020 - 4:30pm
Venue:
Zoom registration required

## ISL colloquium presents "Computational Barriers to Estimation from Low-Degree Polynomials"

Topic:
Computational Barriers to Estimation from Low-Degree Polynomials
Abstract / Description:

One fundamental goal of high-dimensional statistics is to detect or recover planted structure (such as a low-rank matrix) hidden in noisy data. A growing body of work studies low-degree polynomials as a restricted model of computation for such problems. Many leading algorithmic paradigms (such as spectral methods and approximate message passing) can be captured by low-degree polynomials, and thus, lower bounds against low-degree polynomials serve as evidence for computational hardness of statistical problems.

Prior work has studied the power of low-degree polynomials for the detection (i.e. hypothesis testing) task. In this work, we extend these methods to address problems of estimating (i.e. recovering) the planted signal instead of merely detecting its presence. For a large class of "signal plus noise" problems, we give a user-friendly lower bound for the best possible mean squared error achievable by any degree-D polynomial. These are the first results to establish low-degree hardness of recovery problems for which the associated detection problem is easy. As applications, we study the planted submatrix and planted dense subgraph problems, resolving (in the low-degree framework) open problems about the computational complexity of recovery in both cases.

Joint work with Tselil Schramm, available at: https://arxiv.org/abs/2008.02269

Date and Time:
Thursday, October 8, 2020 - 4:30pm
Venue:
Zoom registration required

Topic:
Abstract / Description:

Many experiments ("A/B tests") in online platforms exhibit interference, where an intervention applied to one market participant influences the behavior of another participant. This interference can lead to biased estimates of the treatment effect of the intervention.

In this talk we first focus on such experiments in two-sided platforms where "customers" book "listings". We develop a stochastic market model and associated mean field limit to capture dynamics in such experiments, and use our model to investigate how the performance of different designs and estimators is affected by marketplace interference effects. Platforms typically use two common experimental designs: demand-side ("customer") randomization (CR) and supply-side ("listing") randomization (LR), along with their associated estimators. We show that good experimental design depends on market balance: in highly demand-constrained markets, CR is unbiased, while LR is biased; conversely, in highly supply-constrained markets, LR is unbiased, while CR is biased. We also introduce and study a novel experimental design based on two-sided randomization (TSR) where both customers and listings are randomized to treatment and control. We show that appropriate choices of TSR designs can be unbiased in both extremes of market balance, while yielding relatively low bias in intermediate regimes of market balance. (This is based on joint work with Hannah Li, Inessa Liskovich, and Gabriel Weintraub.)

Time permitting, we will conclude the talk with some discussion of other experimental designs used by online platforms to address interference, including adaptive designs such as switchback experiments, and clustered experimental designs. The goal is to provide an overview of some of the open challenges that arise in this domain. (This part of the talk is based in part on joint work with Peter Glynn and Mohammad Rasouli.)

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Thursday, November 19, 2020 - 4:30pm
Venue:
Zoom

## ISL Colloquium presents "Connecting Online Optimization and Control"

Topic:
Connecting Online Optimization and Control
Abstract / Description:

Online optimization is a powerful framework in machine learning that has seen numerous applications to problems in distributed systems, robotics, autonomous planning, and sustainability. In my group at Caltech, we began by applying online optimization to 'right-size' capacity in data centers a decade ago; and now we have used tools from online optimization to develop algorithms for demand response, energy storage management, video streaming, drone navigation, autonomous driving, and beyond. In this talk, I will highlight both the applications of online optimization and the theoretical progress that has been driven by these applications. Over the past decade, the community has moved from designing algorithms for one-dimensional problems with restrictive assumptions on costs to general results for high-dimensional non-convex problems that highlight the role of constraints, predictions, delay, and more. In the last two years, a connection between online optimization and adversarial control has emerged, and I will highlight how advances in online optimization can lead to advances in the control of linear dynamical systems.

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Thursday, October 22, 2020 - 4:30pm
Venue:
Zoom

## ISL Colloquium presents "Strategies for Active Machine Learning"

Topic:
Strategies for Active Machine Learning
Abstract / Description:

The field of Machine Learning (ML) has advanced considerably in recent years, but mostly in well-defined domains and often using huge amounts of human-labeled training data. Machines can recognize objects in images and translate text, but they must be trained with more images and text than a person can see in nearly a lifetime. The computational complexity of training has been offset by recent technological advances, but the cost of training data is measured in terms of the human effort in labeling data. People are not getting faster nor cheaper, so generating labeled training datasets has become a major bottleneck in ML pipelines.

Active ML aims to address this issue by designing learning algorithms that automatically and adaptively select the most informative examples for labeling so that human time is not wasted labeling irrelevant, redundant, or trivial examples. This talk explores the development of active ML theory and methods over the past decade, including a new approach applicable to kernel methods and neural networks, which views the learning problem through the lens of representer theorems. This perspective highlights the effect of adding a new training example on the functional representation, leading to a new criterion for actively selecting examples.

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Thursday, October 15, 2020 - 4:30pm
Venue:
Zoom

## ISL Colloquium presents "Optimizing the Cost of Distributed Learning"

Topic:
Optimizing the Cost of Distributed Learning
Abstract / Description:

As machine learning models are trained on ever-larger and more complex datasets, it has become standard to distribute this training across multiple physical computing devices. Such an approach offers a number of potential benefits, including reduced training time and storage needs due to parallelization. Distributed stochastic gradient descent (SGD) is a common iterative framework for training machine learning models: in each iteration, local workers compute parameter updates on a local dataset. These are then sent to a central server, which aggregates the local updates and pushes global parameters back to local workers to begin a new iteration. Distributed SGD, however, can be expensive in practice: training a typical deep learning model might require several days and thousands of dollars on commercial cloud platforms. Cloud-based services that allow occasional worker failures (e.g., locating some workers on Amazon spot or Google preemptible instances) can reduce this cost, but may also reduce the training accuracy. We quantify the effect of worker failure and recovery rates on the model accuracy and wall-clock training time, and show both analytically and experimentally that these performance bounds can be used to optimize the SGD worker configurations. In particular, we can optimize the number of workers that utilize spot or preemptible instances. Compared to heuristic worker configuration strategies and standard on-demand instances, we dramatically reduce the cost of training a model, with modest increases in training time and the same level of accuracy. Finally, we discuss implications of our work for federated learning environments, which use a variant of distributed SGD. Two major challenges in federated learning are unpredictable worker failures and a heterogeneous (non-i.i.d.) distribution of data across the workers, and we show that our characterization of distributed SGD's performance under worker failures can be adapted to this setting.

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Thursday, October 1, 2020 - 4:30pm

## ISL Colloquium presents "Quantum Renyi relative entropies and their use"

Topic:
Quantum Renyi relative entropies and their use
Abstract / Description:

The past decade of research in quantum information theory has witnessed extraordinary progress in understanding communication over quantum channels, due in large part to quantum generalizations of the classical Renyi relative entropy. One generalization is known as the sandwiched Renyi relative entropy and finds its use in characterizing asymptotic behavior in quantum hypothesis testing. It has also found use in establishing strong converse theorems (fundamental communication capacity limitations) for a variety of quantum communication tasks. Another generalization is known as the geometric Renyi relative entropy and finds its use in establishing strong converse theorems for feedback assisted protocols, which apply to quantum key distribution and distributed quantum computing scenarios. Finally, a generalization now known as the Petz–Renyi relative entropy plays a critical role for statements of achievability in quantum communication. In this talk, I will review these quantum generalizations of the classical Renyi relative entropy, discuss their relevant information-theoretic properties, and the applications mentioned above.

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Thursday, September 17, 2020 - 4:30pm
Venue:
Zoom

## IT-Forum presents "A Very Sketchy Talk"

Topic:
A Very Sketchy Talk
Abstract / Description:

We give an overview of dimensionality reduction methods, or sketching, for a number of problems in optimization, first surveying work using these methods for classical problems, which gives near optimal algorithms for regression, low rank approximation, and natural variants. We then survey recent work applying sketching to column subset selection, kernel methods, sublinear algorithms for structured matrices, tensors, trace estimation, and so on. The focus in the talk will be on fast algorithms.

The Information Theory Forum (IT-Forum) at Stanford ISL is an interdisciplinary academic forum which focuses on mathematical aspects of information processing. With a primary emphasis on information theory, we also welcome researchers from signal processing, learning and statistical inference, control and optimization to deliver talks at our forum. We also warmly welcome industrial affiliates in the above fields. The forum is typically held every Friday at 1:15 pm during the academic year.

Until further notice, the IT Forum convenes exclusively via Zoom (on Fridays at 1:15pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJwkf-uvqjoqHNIWxY4HHon4K107QMo22PVR to receive the Zoom meeting details via email.

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Friday, November 13, 2020 - 1:15pm
Venue:
Zoom

## ISL Colloquium presents "Federated Learning at Google and Beyond"

Topic:
Federated Learning at Google and Beyond
Abstract / Description:

Federated Learning enables mobile devices to collaboratively learn a shared prediction model or analytic while keeping all the training data on device, decoupling the ability to do machine learning from the need to store the data in the cloud. It embodies the principles of focused collection and data minimization, and can mitigate many of the systemic privacy risks and costs resulting from traditional, centralized machine learning. In this talk, I will discuss: (1) how federated learning differs from more traditional distributed machine learning paradigms, focusing on the main defining characteristics and challenges of the federated learning setting; (2) practical algorithms for federated learning that address the unique challenges of this setting; (3) extensions to federated learning, including differential privacy, secure aggregation, and compression for model updates, and (4) a range of valuable research directions that could have significant real-world impact.

Date and Time:
Friday, February 28, 2020 - 1:15pm
Venue:
Packard 202

## RL forum presents "Temporal Abstraction in Reinforcement Learning with the Successor Representation"

Topic:
Temporal Abstraction in Reinforcement Learning with the Successor Representation
Abstract / Description:

Reasoning at multiple levels of temporal abstraction is one of the key abilities for artificial intelligence. In the reinforcement learning problem, this is often instantiated with the options framework. Options allow agents to make predictions and to operate at different levels of abstraction within an environment. Nevertheless, when a reasonable set of options is not known beforehand, there are no definitive answers for characterizing which options one should consider. Recently, a new paradigm for option discovery has been introduced. This paradigm is based on the successor representation (SR), which defines state generalization in terms of how similar successor states are. In this talk I'll discuss the existing methods from this paradigm, providing a big picture look at how the SR can be used in the options framework. I'll present methods for discovering "bottleneck" options, as well as options that improve an agent's exploration capabilities. I'll also discuss the option keyboard, which uses the SR to extend a finite set of options to a combinatorially large counterpart without additional learning.

Date and Time:
Tuesday, February 25, 2020 - 2:00pm
Venue:
Packard 202

## IT Forum presents "Towards Model Agnostic Robustness"

Topic:
Towards Model Agnostic Robustness
Abstract / Description:

It is now common practice to try and solve machine learning problems by starting with a complex existing model or architecture, and fine-tuning/adapting it to the task at hand. However, outliers, errors or even just sloppiness in training data often lead to drastic drops in performance.

We investigate a simple generic approach to correct for this, motivated by a classic statistical idea: trimmed loss. This advocates jointly (a) selecting which training samples to ignore, and (b) fitting a model on the remaining samples. As such this is computationally infeasible even for linear regression. We propose and study the natural iterative variant that alternates between these two steps (a) and (b) - each of which individually can be easily accomplished in pretty much any statistical setting. We also study the batch-SGD variant of this idea. We demonstrate both theoretically (for generalized linear models) and empirically (for moderate-sized neural network models) that this effectively recovers accuracy in the presence of bad training data.

This work is joint with Yanyao Shen and Vatsal Shah and appears in NeurIPS 2019, ICML 2019 and AISTATS 2020.

Date and Time:
Thursday, October 29, 2020 - 4:30pm
Venue:
Registration required

## IT-Forum presents "A Very Sketchy Talk"

Topic:
A Very Sketchy Talk
Abstract / Description:

We give an overview of dimensionality reduction methods, or sketching, for a number of problems in optimization, first surveying work using these methods for classical problems, which gives near optimal algorithms for regression, low rank approximation, and natural variants. We then survey recent work applying sketching to column subset selection, kernel methods, sublinear algorithms for structured matrices, tensors, trace estimation, and so on. The focus in the talk will be on fast algorithms.

The Information Theory Forum (IT-Forum) at Stanford ISL is an interdisciplinary academic forum which focuses on mathematical aspects of information processing. With a primary emphasis on information theory, we also welcome researchers from signal processing, learning and statistical inference, control and optimization to deliver talks at our forum. We also warmly welcome industrial affiliates in the above fields. The forum is typically held every Friday at 1:15 pm during the academic year.

Until further notice, the IT Forum convenes exclusively via Zoom (on Fridays at 1:15pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJwkf-uvqjoqHNIWxY4HHon4K107QMo22PVR to receive the Zoom meeting details via email.

The ISL Colloquium meets weekly during the academic year. Seminars are each Thursday at 4:30pm PT, unless indicated otherwise.

Until further notice, the ISL Colloquium convenes exclusively via Zoom (on Thursdays at 4:30pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJckfuCurzkvEtKKOBvDCrPv3McapgP6HygJ to receive the Zoom meeting details via email.

Date and Time:
Friday, November 13, 2020 - 1:15pm
Venue:
Zoom

## IT-Forum presents "High-accuracy Optimality and Limitation of the Profile Maximum Likelihood"

Topic:
High-accuracy Optimality and Limitation of the Profile Maximum Likelihood
Abstract / Description:

Symmetric properties of distributions arise in multiple settings, where for each of them separate estimators have been developed. Recently, Orlitsky et al. showed that a single estimator, called the profile maximum likelihood (PML), achieves the optimal sample complexity universally for many properties and any accuracy parameter larger than \$n^{-1/4}\$, where \$n\$ is the sample size. They also raised the question whether this low-accuracy range is an artifact of the analysis or a fundamental limitation of the PML, which remained open after several subsequent work.

In this talk, we provide a complete answer to this question and characterize the tight performance of PML in the high-accuracy regime. On the positive side, we show that the PML remains sample-optimal for any accuracy parameter larger than \$n^{-1/3}\$ using a novel chaining property of the PML distributions. In particular, the PML distribution itself is an optimal estimator of the sorted hidden distribution. On the negative side, we show that the PML as well as any adaptive approach cannot be universally sample-optimal when the accuracy parameter is below \$n^{-1/3}\$, and characterize the exact penalty for adaptation via a matching adaptation lower bound.

Based on joint work with Kirankumar Shiragur. The full papers are available online at https://arxiv.org/abs/2004.03166 and https://arxiv.org/abs/2008.11964.

The Information Theory Forum (IT-Forum) at Stanford ISL is an interdisciplinary academic forum which focuses on mathematical aspects of information processing. With a primary emphasis on information theory, we also welcome researchers from signal processing, learning and statistical inference, control and optimization to deliver talks at our forum. We also warmly welcome industrial affiliates in the above fields. The forum is typically held every Friday at 1:15 pm during the academic year.

Until further notice, the IT Forum convenes exclusively via Zoom (on Fridays at 1:15pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJwkf-uvqjoqHNIWxY4HHon4K107QMo22PVR to receive the Zoom meeting details via email.

Date and Time:
Friday, September 25, 2020 - 1:15pm
Venue:
Zoom registration req'd

## Learning to Bid in Repeated First-price Auctions

Topic:
Learning to Bid in Repeated First-price Auctions
Abstract / Description:

First-price auctions have very recently swept the online advertising industry, replacing second-price auctions as the predominant auction mechanism on many platforms. This shift has brought forth important challenges for a bidder: how should one bid in a first-price auction where it is no longer optimal to bid one's private value truthfully and hard to know the others' bidding behaviors? To answer this question, we study online learning in repeated first-price auctions, and consider various scenarios involving different assumptions on the characteristics of the other bidders' bids, of the bidder's private valuation, of the feedback structure of the auction, and of the reference policies with which our bidder competes. For all of them, we characterize the essentially optimal performance and identify computationally efficient algorithms achieving it. Experimentation on first-price auction datasets from Verizon Media demonstrates the promise of our schemes relative to existing bidding algorithms.

Based on joint work with Aaron Flores, Erik Ordentlich, Tsachy Weissman, and Zhengyuan Zhou. The full papers are available online at https://arxiv.org/abs/2003.09795 and https://arxiv.org/abs/2007.04568

The Information Theory Forum (IT-Forum) at Stanford ISL is an interdisciplinary academic forum which focuses on mathematical aspects of information processing. With a primary emphasis on information theory, we also welcome researchers from signal processing, learning and statistical inference, control and optimization to deliver talks at our forum. We also warmly welcome industrial affiliates in the above fields. The forum is typically held every Friday at 1:15 pm during the academic year.

Until further notice, the IT Forum convenes exclusively via Zoom (on Fridays at 1:15pm PT) due to the ongoing pandemic. To avoid "Zoom-bombing", we ask attendees to input their email address here https://stanford.zoom.us/meeting/register/tJwkf-uvqjoqHNIWxY4HHon4K107QMo22PVR to receive the Zoom meeting details via email.

Date and Time:
Friday, September 18, 2020 - 1:15pm
Venue:
Zoom

## IT Forum presents "Fundamental barriers to estimation in high-dimensions"

Topic:
Fundamental barriers to estimation in high-dimensions
Abstract / Description:

Modern large-scale statistical models require to estimate thousands to millions of parameters. Understanding the tradeoff between statistical optimality and computational tractability in such models remains an outstanding challenge. Under a random design assumption, we establish lower bounds to statistical estimation with two popular classes of tractable estimators in several popular statistical models. First, in high-dimensional linear models we show that a large gap often exists between the optimal statistical error and that achieved by least squares with a convex penalty. Examples of such estimators include the Lasso, ridge regression, and MAP estimation with log-concave priors and Gaussian noise. Second, in generalized linear models and low-rank matrix estimation problems, we introduce the class of 'general first order methods,' examples of which include gradient descent, projected gradient descent, and their accelerated versions. We derive lower bounds for general first order methods which are tight up to asymptotically negligible terms. Our results demonstrate a gap to statistical optimality for general first order methods in both sparse phase retrieval and sparse PCA.

This is joint work with Andrea Montanari and Yuchen Wu.

Date and Time:
Friday, April 3, 2020 - 1:15pm
Venue:
Zoom: stanford.zoom.us/j/516499996

## IT Forum presents "What Hockey Teams and Foraging Animals Can Teach Us About Feedback Communication"

Topic:
What Hockey Teams and Foraging Animals Can Teach Us About Feedback Communication
Abstract / Description:

Suppose we wish to communicate over a memoryless channel with known statistics. How can the use of a feedback link from the receiver to the transmitter help? We introduce a novel mechanism for using feedback, called timid/bold coding, and we show that for some channels timid/bold coding yields a strict asymptotic improvement over the best non- feedback schemes. We also show that for a broad class of channels, feedback is useful if and only if timid/bold coding is applicable. The talk contains a puzzle (recently featured on the FiveThirtyEight website), a life lesson, a (potential) practical application, and some stochastic calculus.

This is joint work with Nirmal Shende and Yucel Altug.

Date and Time:
Friday, March 6, 2020 - 1:15pm
Venue:
Packard 202

## RL forum presents "Temporal Abstraction in Reinforcement Learning with the Successor Representation"

Topic:
Temporal Abstraction in Reinforcement Learning with the Successor Representation
Abstract / Description:

Reasoning at multiple levels of temporal abstraction is one of the key abilities for artificial intelligence. In the reinforcement learning problem, this is often instantiated with the options framework. Options allow agents to make predictions and to operate at different levels of abstraction within an environment. Nevertheless, when a reasonable set of options is not known beforehand, there are no definitive answers for characterizing which options one should consider. Recently, a new paradigm for option discovery has been introduced. This paradigm is based on the successor representation (SR), which defines state generalization in terms of how similar successor states are. In this talk I'll discuss the existing methods from this paradigm, providing a big picture look at how the SR can be used in the options framework. I'll present methods for discovering "bottleneck" options, as well as options that improve an agent's exploration capabilities. I'll also discuss the option keyboard, which uses the SR to extend a finite set of options to a combinatorially large counterpart without additional learning.

Date and Time:
Tuesday, February 25, 2020 - 2:00pm
Venue:
Packard 202

## IT Forum presents "Why should information theorists and probabilists care about blockchains?"

Topic:
Why should information theorists and probabilists care about blockchains?
Abstract / Description:

The invention of blockchains in 2008 opened up for the first time the possibility of large scale decentralized trust systems. In the past few years, the design and analysis of blockchains have received significant attention from cryptography, security and distributed systems communities. In this talk, I argue that there are many interesting problems for information theorists and probabilists as well. I will discuss two particular classes of problems: 1) security analysis of blockchains as convergence analysis of random tree processes; 2) the use of coding to efficiently scale blockchains. I will also briefly discuss a course I plan to teach in the Spring on the topic.

Part of this talk is on joint work with Amir Dembo and Ofer Zeitouni.

Date and Time:
Friday, January 31, 2020 - 1:00pm
Venue:
Packard 202

## Q-Farm Quantum Seminar Series presents "Surprises from Time Crystals"

Topic:
Surprises from Time Crystals
Abstract / Description:

Time crystals are new states of matter that only exist in an out-of-equilibrium setting. I will review the state of this rapidly evolving field, focusing in particular on some of the remarkable properties of this phase, and the surprises coming out of its study. I will provide a detailed overview of existing experiments, with a view towards identifying the ingredients needed for an unambiguous observation of this phase in the future.

Date and Time:
Wednesday, January 29, 2020 - 12:00pm
Venue:
Hansen Physics & Astrophysics Building, 102/103

## ISL & IT-Forum present "Approaching Capacity at Short Blocklengths"

Topic:
Approaching Capacity at Short Blockengths
Abstract / Description:

This talk explores a family of recent results directed to approaching capacity at short blocklengths on the order of 50-500 channel transmissions. Convolutional codes out-perform polar codes and LDPC codes to approach the random coding union bound with low complexity when used with an optimized CRCs and list decoding. This perspective rehabilitates "catastrophic" convolutional codes, which are more properly understood for finite blocklengths as clever expurgation rather than any sort of catastrophe. This approach also provides a low-complexity approach for maximum-likelihood decoding of high-rate BCH codes. The use of variable length coding, i.e. incremental redundancy controlled with simple ACK/NACK feedback, allows capacity to be closely approached by practical codes with fewer than 500 channel uses. This talk reviews the information-theoretic results of Polyanskiy with respect to ACK/NACK feedback, presents new results extending the classic approach of Horstein for full feedback, and shows how to optimize the number and length of incremental redundancy transmissions (and feedback transmissions) for a variable-length code with feedback (i.e. a type-II hybrid ARQ). The talk also shows how to avoid entirely the overhead of a CRC in a hybrid ARQ setting by directly computing the reliability of convolutional codeword decisions. Finally, attendees will learn about a novel communications architecture that allows the use of incremental redundancy even without feedback.

Date and Time:
Friday, January 24, 2020 - 1:15pm
Venue:
Packard 202

## OSA/SPIE present "Silicon-based Optical Biosensors"

Topic:
Silicon-based Optical Biosensors
Abstract / Description:

Silicon-based optical biosensors hold great promise as low-cost, lab-on-chip sensor array elements due to their compatibility with both standard microelectronics processing and standard surface functionalization techniques. Key metrics of these optical biosensors include sensitivity, which is largely derived from the level of interaction between light and the target molecules to be detected, selectivity, and robustness. This talk will present several approaches to increase detection sensitivity through the use of on-chip resonant cavity geometries and porous substrates. The outlook for translating promising silicon-based benchtop sensing platforms into reliable clinical and point-of-care diagnostic devices will also be discussed.

Date and Time:
Thursday, December 3, 2020 - 1:30pm
Venue:

## AP 483 Seminar presents "Organic small molecule integrated photonics"

Topic:
Organic small molecule integrated photonics
Abstract / Description:

The initial, landmark integrated photonic devices relied on silicon and III-V materials, and recent advances in material fabrication and deposition methods have enabled a plethora of new technologies based on materials with higher optical nonlinearities, including 2D materials and organic polymers. However, nonlinear optical (NLO) organic small molecules have not experienced similar growth due to a perceived environmental instability and to challenges related to intra and intermolecular interactions. Because NLO small molecules have NLO coefficients that are orders of magnitude larger than conventional optical materials, developing strategies to fabricate optical devices could enable significant performance improvements. In recent work, we combined conventional top-down fabrication methods with bottom-up techniques to develop on-chip devices that incorporated NLO optical small molecules. These hybrid systems provide access to optical behavior and performance not attainable with conventional material systems. In this seminar, I will discuss a couple examples of NLO small molecule integrated resonators, including Raman lasers and all optically-switchable devices.

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Date and Time:
Monday, October 19, 2020 - 4:15pm

## AP 483 seminar presents "Sensing beyond the conventional limits: vortices, viscosity and molecular vibrations"

Topic:
Sensing beyond the conventional limits: vortices, viscosity and molecular vibrations
Abstract / Description:

It has been known for over a century that, in quantum physics, even the act of looking can have dramatic consequences. For instance, it kills the cat in Schrödinger’s famous thought experiment. However, it has proved extremely difficult to reach regimes where such effects play a role, let alone to use them as a tool to enhance measurement technologies.

Over the past decade, however, advances in photonics and nanotechnology have allowed us to engineer both devices and states of light which exhibit this distinctive quantum behavior [1]. These “quantum optomechanical devices” consist of a nanoscale mechanical object – for example, a nanoparticle, molecule or cantilever – coupled to light via radiation pressure, often concentrated in a tiny optical cavity. In essence, they are miniature versions of the kilometer-scale interferometers that have enabled the extraordinary detection of gravitational waves from distant black hole collisions. Quite remarkably, they can allow measurements of motion at the sub-attometre level – more than a thousand times below the width of an atomic nucleus. At a fundamental level, this allows us to ask new questions of quantum physics for macroscopic systems consisting of trillions of atoms. It also provides a way to build precision optical sensors that far outperform the current state-of-the-art.

In this talk, I will provide an overview of optomechanical sensors developed in my laboratory, with a particular focus on applications in the bioscience and in studying superfluid helium, a strongly interacting quantum liquid. This includes the observation of coherent vortex dynamics in two-dimensional superfluid helium [2]; the engineering of extreme Brillouin nonlinearities in superfluid thin-films [3]; a quantum-light microscope that provides absolute quantum advantage, allowing the observation of molecular vibrations of nanoscale biological structures that would otherwise be unresolvable [4]; and optical tweezers that allow tracking of the instantaneous

velocity of trapped particles, and through this orders-of-magnitude faster measurements of cell properties reaching microsecond timescales [5].

1. Bowen and Milburn, Quantum Optomechanics, CRC Press (2016).
2. Science 366 1480 (2019).
3. Nature Physics 16 417 (2020); also Nature Physics 12 788 (2016).
4. arXiv:2004.00178 (2020).
5. arXiv:2007.03066 (2020); also Nature Photonics 11, 477-481 (2017); 9, 669-673 (2015); 7, 229-233 (2013)

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Organized by Prof. Amir Safavi-Naeini

Date and Time:
Monday, September 21, 2020 - 4:15pm
Venue:
Zoom

## AP 483 seminar presents "Quantum computing over the rainbow: the quantum optical frequency comb as a platform for measurement-based universal quantum computing"

Topic:
Quantum computing over the rainbow: the quantum optical frequency comb as a platform for measurement-based universal quantum computing
Abstract / Description:

An ultrafast laser emits vastly multimode light over a broad spectral band, a.k.a. the optical frequency comb (OFC), but the emission happens but one photon at a time, if in a stimulated manner, and no entanglement is created in the light. Changing the gain medium from linear (one-photon) to nonlinear (two-photon) yields an optical parametric oscillator which features massively multipartite entanglement of the OFC modes, as demonstrated experimentally by our group and others. This entanglement can then be exquisitely tailored to cluster states with specific graphs, in particular the two-dimensional ones that are universal for measurement-based, one-way quantum computing. It is worth noting that this requires only sparse experimental resources that are highly compatible with integrated optics, thereby paving the way to the realization of practical, fault-tolerant quantum computers.

This seminar is sponsored by the department of Applied Physics and the Ginzton Laboratory.

Organized by Prof. Amir Safavi-Naeini

Date and Time:
Monday, September 14, 2020 - 4:15pm
Venue:
Zoom

## OSA Color Technical Group presents "Modeling the Initial Steps of Human Vision"

Topic:
Modeling the Initial Steps of Human Vision
Abstract / Description:

Vision guides thought and action. To do so usefully it must inform us about critical features of the world around us. What we can learn about the world is limited by the initial stages of visual processing. Physicists, biologists and psychologists have created quantitative models of these stages, and these models enable us to quantify the encoded information. We have integrated these models as image computable software: the Image Systems Engineering Toolbox for Biology (ISETBio). The software is an extensible set of open-source modules that model the three-dimensional scene spectral radiance, retinal image formation (physiological optics), spatial sampling by the cone photoreceptor mosaic, fixational eye movements, and phototransduction. This webinar, hosted by the OSA Color Technical Group, will provide an overview of the ISETBio modules as well as examples of how to use the software to understand and model human visual performance.

Hosted By: Color Technical Group

Date and Time:
Tuesday, July 21, 2020 - 9:00am

## OSA/SPIE, SPRC and Ginzton Lab present "Workshop on Inverse Design using SPINS"

Topic:
Workshop on Inverse Design using SPINS
Abstract / Description:

The goal of this workshop is to provide attendees with a practitioner's perspective of the basic ingredients and tools of inverse design using SPINS [1], a photonic optimization framework developed at Jelena Vuckovic's Nanoscale and Quantum Photonics Lab over the past decade.
SPINS is a flexible inverse-design platform that is compatible with a variety of device parametrizations, electromagnetic solvers, and objective functions. This has enabled SPINS to be used to design and experimentally demonstrate functional devices in a wide variety of different application areas including grating couplers [2-4], optical routing for LiDAR [5], and electron accelerators [6].
In this workshop, we will work through examples of using SPINS to design simple devices, including grating couplers and silicon photonics devices. We will emphasize how attendees can adapt these examples for their own applications and discuss some practical considerations for using inverse-design to produce functional devices. All example code will be available online after the workshop.

[1] Su, L. et al. Nanophotonic inverse design with SPINS: software architecture and practical considerations. Appl. Phys. Rev. https://doi.org/10.1063/1.5131263 (2020).
[2] Su, L. et al. Fully-automated optimization of grating couplers. Optics Express, 26(4): 4023–4034, 2018.
[3] Sapra, N. V. et al. Inverse design and demonstration of broadband grating couplers. IEEE J. Sel. Top. Quantum Electron. 25, 1–7 (2019).
[4] Dory, etl al. Inverse-designed diamond photonics. Nature Communications, 10(1):3309, 2019.
[5] Yang, K. Skarda, J. et al., Inverse-designed non-reciprocal pulse router for chip-based LiDAR, Nature Photonics DOI: 10.1038/s41566-020-0606-0 (2020)
[6] Sapra, N. V. et al., On-chip integrated laser-driven particle accelerator. Science 367, 79-83 (2020).

Date and Time:
Thursday, June 4, 2020 - 1:30pm
Venue:
Zoom ID: 945 5728 7546 (password required)

## OSA/SPIE, SPRC and Ginzton Lab present "Next Generation Photonics"

Topic:
Next Generation Photonics
Abstract / Description:

Over the past few decades, silicon photonics has revolutionized photonic integrated circuits by leveraging the semiconductor CMOS manufacturing infrastructure for low cost, high performance devices and systems. However, key fundamental challenges of the field remain unsolved: packaging the devices with optical fibers and generating light on chip.

We developed a novel approach for fiber packaging based on fusing the fiber and chip together. Connecting a silicon photonic chip across long distances requires attaching optical fibers to the chip. In practice, the packaging of optical fibers to photonic devices is time consuming, lossy, and expensive. This process is usually done by gluing the fiber and chip together using optical adhesives. By fusion splicing the chip and fiber together we have demonstrated losses as low as 1dB for single fibers and 2.5dB for an array of four fibers.

Imagine a laser as thick as an atom that is compatible with silicon photonics. Silicon based materials are passive. Since silicon is an indirect band gap material it is a poor light emitter. To generate light on a silicon photonic integrated circuit, you need to integrate active materials, which are usually not compatible with CMOS, with the device. Two-dimensional materials are excellent candidates for light sources, modulators, and detectors; and they are compatible with CMOS electronic manufacturing in the back end. We recently demonstrated the first fully on-chip 2D laser. Due to their thickness and transfer process, we envision electronic-photonic devices with many optical layers, each with their own lasers, modulators, and detectors based on 2D materials. Our recent demonstration completes the set of active devices completely based on 2D materials.

Date and Time:
Wednesday, May 6, 2020 - 1:00pm
Venue:
Zoom ID: 991 2584 9377

## OSA/SPIE with SPRC and Ginzton Lab present "Next Generation Photonics"

Topic:
Next Generation Photonics
Abstract / Description:

Over the past few decades, silicon photonics has revolutionized photonic integrated circuits by leveraging the semiconductor CMOS manufacturing infrastructure for low cost, high performance devices and systems. However, key fundamental challenges of the field remain unsolved: packaging the devices with optical fibers and generating light on chip.

We developed a novel approach for fiber packaging based on fusing the fiber and chip together. Connecting a silicon photonic chip across long distances requires attaching optical fibers to the chip. In practice, the packaging of optical fibers to photonic devices is time consuming, lossy, and expensive. This process is usually done by gluing the fiber and chip together using optical adhesives. By fusion splicing the chip and fiber together we have demonstrated losses as low as 1dB for single fibers and 2.5dB for an array of four fibers.

Imagine a laser as thick as an atom that is compatible with silicon photonics. Silicon based materials are passive. Since silicon is an indirect band gap material it is a poor light emitter. To generate light on a silicon photonic integrated circuit, you need to integrate active materials, which are usually not compatible with CMOS, with the device. Two-dimensional materials are excellent candidates for light sources, modulators, and detectors; and they are compatible with CMOS electronic manufacturing in the back end. We recently demonstrated the first fully on-chip 2D laser. Due to their thickness and transfer process, we envision electronic-photonic devices with many optical layers, each with their own lasers, modulators, and detectors based on 2D materials. Our recent demonstration completes the set of active devices completely based on 2D materials.

Date and Time:
Monday, May 4, 2020 - 12:05pm
Venue:
Zoom ID: 991 2584 9377 (contact organizers for password)

## Stanford Photonics Research Center (SPRC) SPECIAL SEMINAR on Global Environmental Measurement and Monitoring Technologies

Topic:
The Source and Fate of Fossil Fuel CO2 Emissions
Abstract / Description:

Fossil fuel burning is the primary driver of increasing atmospheric carbon dioxide (CO2), which recently passed the landmark 400 parts per million mark. Most nations have committed to reducing their emissions under the Paris Agreement and many sub-national entities have made similar commitments. To ensure that the promised emission reductions are achieved, an understanding of emission rates from fossil fuel point sources, cities, regions and countries, and of uptake of carbon into the land and oceans, is critical.

Traditionally, "bottom-up" economic information has been used to determine fossil fuel emission rates, and biomass estimates are used to evaluate the land carbon sink. In this presentation, I will discuss the use of a "top-down" approach that uses atmospheric measurements and modelling that complements the bottom-up method. In particular, I will focus on the use of radiocarbon (14C) in CO2 as a tracer for the fossil fuel component of atmospheric CO2. I will showcase how the method can be used to quantify fossil fuel CO2 emissions and land carbon uptake, giving examples from USA and New Zealand.

Date and Time:
Thursday, March 12, 2020 - 3:30pm
Venue:
Spilker 232

## OSA/SPIE, SPRC and Ginzton Lab present "Applications of Precision Motion Control in the Optics and Photonics Industry"

Topic:
Applications of Precision Motion Control in the Optics and Photonics Industry
Abstract / Description:

Precision Automation has become a rapidly growing industry as manufacturers must develop ways to achieve ever-tightening tolerances. Likewise, research must push the envelope in order to maintain the increasingly advancing pace of technology. Aerotech has been providing assistance and solutions to researchers within academia and laboratories worldwide for 50 years and continues to empower students and technicians to complete and publish their research. In this seminar, we will discuss why automation and motion control is important as well as the various methods by which Aerotech has helped expand the boundaries of research in the field of optics and photonics. The topics we will cover in this session include:

• What Precision Motion Control is and why it is important
• How to choose the right set of motion control
• How Aerotech has helped further research in:
• Optics design and manufacturing
• Light sources and beamlines
• Optical positioning and ultra-fast spectroscopy
• Fiber alignment
• Laser Processing

After the presentation, a round-table discussion will be held during which we'd like to hear from you! Please bring with you any topics and - more importantly - any tough issues that you'd like to pose pertaining to automated motion control within your specific research field.

Date and Time:
Tuesday, March 3, 2020 - 4:15pm
Venue:
Spilker 232

## SCIEN and EE292E present "Low Light High Dynamic Range CMOS Imaging Sensors"

Topic:
Low Light High Dynamic Range CMOS Imaging Sensors
Abstract / Description:

CMOS image sensors (CIS) are widely used in commercial, industrial and scientific applications. The low light sensors are useful for day-through-night imaging, capturing images in extreme lighting conditions from daylight through quarter-moon darkness. The low light sensors are required for numerous applications like surveillance, security, and mobile applications. Low illumination imaging performance and dynamic range (DR) of an image sensor are limited by the temporal noise of the photodetector and readout signal chain. I will overview some of our work in the domain of low light imaging, starting from a novel pixels increasing the conversion gain and simultaneously reducing the temporal noise of the pixels to more advanced pixel architectures capable of lowering the temporal noise below sub electron level. The dual effect of increased conversion gain and reduced noise, increases the effective dynamic range in the low illumination conditions and provide an alternative to commercial low light sensors like image intensifiers, electron bombardment, EMCMOS. The use of these high responsive light sensors in applications like surveillance, security, reconnaissance UAVs, bio-luminescence and imaging systems in space platform will be discussed.

Register in advance for this meeting:
Date and Time:
Wednesday, December 9, 2020 - 9:30am
Venue:

## SCIEN and EE292E presents "Immersive Light Field Video with a Layered Mesh Representation"

Topic:
Immersive Light Field Video with a Layered Mesh Representation
Abstract / Description:

In this talk I will describe our system for capturing, reconstructing, compressing, and rendering high quality immersive light field video. We record immersive light fields using a custom array of 46 time-synchronized cameras distributed on the surface of a hemispherical, 92cm diameter dome. From this data we produce 6DOF volumetric videos with a wide 80-cm viewing baseline, 10 pixels per degree angular resolution, and a wide field of view (>220 degrees), at 30fps video frame rates. Even though the cameras are placed 18cm apart on average, our system can reconstruct objects as close as 20cm to the camera rig. We accomplish this by leveraging the recently introduced DeepView view interpolation algorithm, replacing its underlying multi-plane image (MPI) scene representation with a collection of spherical shells which are better suited for representing panoramic light field content. We further process this data to reduce the large number of shell layers to a small, fixed number of RGBA+depth layers without significant loss in visual quality. The resulting RGB, alpha, and depth channels in these layers are then compressed using conventional texture atlasing and video compression techniques. The final, compressed representation is lightweight and can be rendered on mobile VR/AR platforms or in a web browser.

Date and Time:
Wednesday, November 18, 2020 - 4:30pm

## SCIEN presents "Label-free optical imaging of living biological systems"

Topic:
Label-free optical imaging of living biological systems
Abstract / Description:

Label-free optical imaging of living biological systems offers rich information that can be of immense value for a variety of biomedical tasks. Despite the exceptional theoretical potential, current label-free microscopy platforms are challenging for real-world clinical and biological applications. The major obstacles include the lack of flexible laser sources, limited contrast, and the challenge of acquiring and interpreting the high-dimensional dataset.

In this talk, I will present new optical imaging platforms and methodologies that will address these challenges. By generating and tailoring coherent supercontinuum from photonic crystal fibers, simultaneous metabolic and structural imaging can be achieved without aids of stains, enabling perturbation-free exploration of living systems. These capabilities further motivate development of analytical tools for image-based segmentation and diagnosis, showing broad potential of this label-free imaging technology in discovering new metabolic biomarkers and enabling real-time point-of-procedure applications.

After registering, you will receive a confirmation email containing information about joining the meeting.

Date and Time:
Wednesday, November 11, 2020 - 4:30pm

## SCIEN and EE292E present "Functional imaging and control of retinal ganglion cells in the living primate eye"

Topic:
Functional imaging and control of retinal ganglion cells in the living primate eye
Abstract / Description:

The encapsulation of the retina inside the eye has always challenged our ability to study the anatomy and physiology of retinal neurons in their native state. Our group is developing new tools using adaptive optics that allow not only structural imaging but also functional recording and control of retinal neurons at a cellular spatial scale. By combining adaptive optics with calcium imaging, we can optically record from hundreds of ganglion cells in the nonhuman primate eye over periods as along as years. This approach is especially well-suited for recording from cells serving the central fovea, which has been difficult to access with microelectrodes. Using optogenetics, we can also directly excite these same ganglion cells with light in the living animal. These capabilities together establish a two-way communication link with retinal ganglion cells. I will discuss the advantages and the current limitations of these approaches, as well as speculate about possible future applications for vision restoration and understanding the role of the retina in perception.

Date and Time:
Wednesday, October 28, 2020 - 4:30pm
Venue:
Registration required

## SCIEN and EE292E present " The Moon Camera"

Topic:
The Moon Camera
Abstract / Description:

My attempts to photograph the Earth from space using the moon as a camera, and several computational imaging projects resulting from those attempts.

Date and Time:
Wednesday, October 14, 2020 - 4:30pm
Venue:

## SCIEN and EE292E present "Computational Imaging, One Photon at a Time"

Topic:
Computational Imaging, One Photon at a Time
Abstract / Description:

Single-photon avalanche diodes (SPADs) are an emerging sensor technology capable of detecting and time-tagging individual photons with picosecond precision. Despite (or perhaps, due to) these capabilities, SPADs are considered specialized devices suitable only for photon-starved scenarios, and restricted to a limited set of niche applications. This raises the following questions: Can SPADs operate not just in low light, but in bright scenes as well? Can SPADs be used not just with precisely controlled active light sources such as pulsed lasers, but under passive, uncontrolled illumination like cellphone or machine vision cameras?
I will describe our recent work on designing computational imaging techniques that (a) enable single-photon sensors to operate across the entire gamut of imaging conditions including high-flux scenes, and (b) leverages SPADs as passive imaging devices for ultra-low light photography. The overall goal is to transform SPADs into all-weather, general-purpose sensors capable of both active and passive imaging, across photon-starved and photon-flooded environments.

Date and Time:
Wednesday, October 7, 2020 - 4:30pm
Venue:
Registration required

## SCIEN and EE292E present "Transforming hearing aids into multisensory perceptual augmentation and health monitoring devices"

Topic:
Transforming hearing aids into multisensory perceptual augmentation and health monitoring devices
Abstract / Description:

With over 466 million people suffering from disabling hearing loss globally according to the World Health Organization, and the number expected to rise to 900 million people by 2050, hearing aids are crucially important medical wearable devices. Untreated hearing loss has been linked to increased risks of social isolation, depression, dementia, fall injuries, and other health issues. In this talk, we will present a new class of in-ear devices with embedded sensors and artificial intelligence, which are shaped to fit an individual with 3D-imaging of the ear geometry. In addition to providing frequency-dependent amplification of sound to compensate for hearing loss, these devices serve as a continuous monitor for important physiological parameters, an automatic fall detection and alert system, as well as a personal assistant with connectivity to the cloud. Furthermore, Bluetooth-paired with a vision aid, these devices present exciting possibilities for multisensory perceptual augmentation of hearing, balance, vision, and memory, helping people live better and more productive lives.

Date and Time:
Wednesday, September 30, 2020 - 4:30pm

## SCIEN and EE292E present "Retinal topography using stripe illumination in a fundus camera"

Topic:
Retinal topography using stripe illumination in a fundus camera
Abstract / Description:

Retinal topography is affected by pathology such as drusen and tumors, and it may be useful to determine topography with fundus imaging when three-dimensional imaging is not available. In this talk, I will present a novel method of retinal topography scanning using the stripe projection technology of the CLARUSTM 700 (ZEISS, Dublin, CA) wide-field fundus camera. The camera projects stripes onto the retina and records images of the returned light while maintaining a small angle between illumination and imaging. We make use of this structured illumination, analyzing neighboring stripes to determine depth -i.e. the retinal topography – from both relative defocus and stripe displacements. The resulting topography maps are finally compared to three-dimensional data from optical coherence tomography imaging.

Date and Time:
Wednesday, September 23, 2020 - 4:30pm
Venue:
Zoom registration required

## SCIEN and EE292E present "Time of flight imaging with single photon detectors”

Topic:
Time of flight imaging with single photon detectors
Abstract / Description:

Single Photon Avalanche Detectors (SPADs) can detect single photon arrival events and in doing so, record a click that can be used to also determine the photon arrival time on the detector with picosecond temporal resolution. This timing capability is finding many uses such as LIDAR and fluorescence lifetime imaging and provides an opportunity for revisiting fundamental imaging concepts by combining SPAD data with computational image retrieval techniques. The computational techniques can in general resort to inverse retrieval approaches or machine learning, with the choice depending on the specific nature of the data and imaging problem at hand. I will overview some of our work, starting from the first attempts to capture light-in-flight using SPAD cameras and covering the topics of non-line-of-sight imaging, imaging through diffusion, extraction of 3D images from time-of-light data only, fluorescence lifetime imaging and coincidence counting for quantum imaging applications.

Date and Time:
Thursday, August 27, 2020 - 10:00am
Venue:
Zoom

## SCIEN and EE292E present "Breaking the Resolution & Speed Limit – Next Generation 3D Printing Technology based on Digital Holography and Temporal Focusing"

Topic:
Breaking the Resolution & Speed Limit – Next Generation 3D Printing Technology based on Digital Holography and Temporal Focusing
Abstract / Description:

Additive manufacturing printing, i.e., 3-D printing, is one of the most important technological innovations in the past few decades. Among the various techniques, two-photon polymerization (TPP) is the most precise 3-D printing process that has been used to create many complex structures for advanced photonic and nanoscale applications, e.g., microrobots, optical memories, metamaterials, photonic crystals, and bio-scaffolds etc. However, to date the technology still remains a laboratory tool due to its high operation cost and limited fabrication rate, i.e., serial laser scanning process. In this seminar, I will present our recent work on the parallelization of the TPP process based on (1) temporal focusing and (2) binary holography, where programmable femtosecond light sheets or tens to hundreds of shaped laser beams are used to substantially improve the rate without sacrificing resolution. In addition, the engineered laser foci can improve the strength and structural integrity of the printed structures. Our experiments demonstrate arbitrarily complex structures can be fabricated at a record-breaking resolution and speed, i.e., lateral/axial resolution: 140 nm/175 nm at 10s mm3/min, which is 3-4 orders of magnitude higher than any existing fabrication methods. Our new methods provide an effective and low-cost solution to scale-up the fabrication of functional micro- and nano-structures (~\$1.5/mm3). This means our technology may play a large role in fields such as healthcare, clean energy and water, computing, and telecommunications.

Date and Time:
Wednesday, August 19, 2020 - 4:30pm
Venue:
Zoom

## Smart Grid Seminar: Dynamic Operation and Lifetime Valuation for Batteries in Markets

Topic:
Dynamic Operation and Lifetime Valuation for Batteries in Markets
Abstract / Description:

This talk introduces a generalized framework for valuating batteries based on their degradation mechanism and market signals. The proposed framework follows a dynamic programming approach, and novel solution algorithms are designed to accurately model years of battery degradation accumulation, while simulating high granularity optimal control for batteries over time resolutions of minutes or even seconds, including the capability to model stochastic operations. Results from this research will provide pricing and operation references for new and used batteries in grid-interactive applications, and demonstrate how batteries of different degradation mechanisms should be priced differently.

Date and Time:
Thursday, November 19, 2020 - 1:30pm
Venue:
Zoom

## Smart Grid Seminar: IoT, Smart Microgrid

Topic:
IoT, Smart Microgrid
Abstract / Description:

Transforming the grid to a Smart Grid requires real time sensing, of energy producers, consumers, stored capacity in batteries, weather conditions, and more, essential to develop usage models, and anticipate need. Along with the ability to take 'action' such as bringing online cleaner sources of power and curbing demand we can help smooth load peaks. In this talk we introduce the open source project EdgeX Foundry and its viability to deliver a secure, extensible, endpoint to collect real time data, develop models, and control devices. Data processing at the edge while reducing network bandwidth needs, reduces response latency, and helps preserve data privacy. We illustrate this in the context of VMware's on-campus electric vehicle charging facility. We conclude by sharing our longer term goals of developing and controlling hierarchical composite device models and introduce open source project Kinney.

Seminars are open to all Stanford students, faculty, staff and community. Register via the RSVP link

Date and Time:
Thursday, October 15, 2020 - 1:30pm
Venue:
Zoom

## Smart Grid Seminar: Increasing Electricity Market Efficiency with Flexible Mechanism

Topic:
Increasing Electricity Market Efficiency with Flexible Mechanism
Abstract / Description:

Current wholesale electricity market designs of day-ahead and real-time markets have inherent inefficiencies that are amplified with greater shares from intermittent renewable sources. Building in flexibility through time and geographically substitutable bids and stochastic unit commitments can yield significant gains. This talk will discuss how to model and assess these advantages and their implications for renewable generation.

Date and Time:
Thursday, October 1, 2020 - 1:30pm
Venue:
Zoom

## SmartGrid Seminar presents "Cyber Security Research: A Data Scientist’s Perspective"

Topic:
Cyber Security Research: A Data Scientist’s Perspective
Abstract / Description:

Situational awareness of computer networks presents many challenges including but not limited to the volume of the data and the dynamic and evolving nature of the problem space. For example, at the perimeter of a corporate enterprise computer network, it is common to see terabytes of network traffic each day, containing millions of unique IP addresses and connection records that number in the hundreds of millions. Celeste will provide an overview and discuss recent trends facing computer security researchers and practitioners. She will describe recent work at Lawrence Livermore National Laboratory to enable analysis of computer networks and encourage an interactive discussion to foster new ideas to address these challenges.

Date and Time:
Thursday, May 7, 2020 - 1:30pm
Venue:
Register at tinyurl.com/Spring-SGS

## Smart Grid Seminar presents "California Microgrids: Lessons Learned"

Topic:
California Microgrids: Lessons Learned
Abstract / Description:

There has been substantial discussion of the use of microgrids as a result of the recent wildfires and Public Safety Power Shutoffs. The ability of microgrids to be considered for these applications now, is in part a result of over a decade of research funded through the California Energy Commission's R&D programs. To date the CEC has issued over \$100M in grants that have funded over 30 microgrids. These microgrids support a variety of applications, with a diversity of ownership models. This presentation will provide an overview of the microgrid research being funded by the CEC and future research areas of interest.

Date and Time:
Thursday, April 23, 2020 - 1:30pm
Venue:

## Smart Grid Seminar presents "Grid Resilience"

Topic:
Grid Resilience
Abstract / Description:

The term "resilience" refers to the ability to prepare for and adapt to changing conditions and withstand and recover rapidly from disruptions. The value of resilience to the electricity system is in enabling the function of other interdependent infrastructure systems and maintaining critical energy-based services. Many resilience approaches are being considered. This webinar will outline some of these perspectives and provide an overview of some of the related ongoing activities within the Department of Energy.

Date and Time:
Thursday, May 21, 2020 - 1:30pm
Venue:

## Smart Grid Seminar presents "Cyber Security Roadmap and Challenge"

Topic:
Abstract / Description:

Cyber security has become a critical priority for electric utilities, which are increasingly dependent on information technology and telecommunication infrastructure to ensure the reliability and security of the electric grid. Additionally, rapid, disruptive changes are happening in electric grids around the world. In many states and countries, initiatives are underway to integrate small, renewable generation into the distribution grid, to meet the local demand for electricity, while reducing the dependency on large, central generation facilities and long-distance transmission. This transition requires new technologies, connectivity, and intelligence, which inherently changes the attack surface of the grid and introduces new stakeholders in the delivery and use of power. This talk provides an introduction to these new cyber security issues, discusses EPRI's approach to addressing the challenges, and presents future research opportunities.

Date and Time:
Thursday, May 28, 2020 - 1:30pm
Venue:

## Smart Grid Seminar presents "Dynamic Electricity Pricing"

Topic:
Dynamic Electricity Pricing
Abstract / Description:

With the rapid growth in residential smart meters across the United States in recent years, most homes will soon be capable of moving to time varying prices for electricity. We discuss a methodology for studying the welfare impacts of different pricing strategies on an electricity market when homes deploy smart, price responsive appliances with forward-looking capabilities, and understanding its effects on social welfare and consumer surplus. We show conditions under which asymptotically, as the number of homes increases, social welfare-maximizing price schedules in equilibrium are linear in load. Real data are used to calibrate a smart thermostat model to compare dynamic pricing strategy against flat and peak pricing strategies when smart thermostats are deployed across ComEd's service territory of 3.5 million residential homes.

Date and Time:
Thursday, May 14, 2020 - 1:30pm
Venue:

## Opportunities in Solar Research and Entrepreneurship

Topic:
Opportunities in Solar Research and Entrepreneurship
Abstract / Description:

Renewable energy generation, storage and energy efficiency are being rapidly integrated into our digitized buildings, transportation system and electric grid. Exponentially expanding opportunities for entrepreneurs to commercialize new products are blossoming. This talk will highlight examples of the wide-ranging resources available to help start-up initiatives accelerate cycles of learning and bring innovations to market. Some specific resources to be reviewed include:

• Stanford's TomKat Center for Sustainable Energy Innovation Transfer
• The NSF I-Corps Program adapted from Steve Blank's Stanford Lean LaunchPad class
• The Elemental Accelerator in East Palo Alto
• Emily Kirsch's PowerHouse Incubator and PowerHouse Ventures in Oakland
• Cyclotron Road at Lawrence Berkeley National Laboratory
• The National Renewable Energy Laboratory Industry Growth Forum and Innovation Incubator
• The Department of Energy Solar Prize
• Dorothy Jones-Davis' Nation of Makers
Date and Time:
Thursday, April 23, 2020 - 2:30pm
Venue:
Zoom

## Bits and Watts presents "Battery Electric Vehicles, Energy Management, and the Smart Grid: An OEM Perspective"

Topic:
Battery Electric Vehicles, Energy Management, and the Smart Grid: An OEM Perspective
Abstract / Description:

Ford is pursuing an aggressive electrification of their product portfolio that includes the launch of the all-electric 2021 Mustang Mach-E, all-electric F-150, and all-electric global commercial van. With 100 kWh of electrical capacity, and potentially more, on board, Ford customers will have enough energy to meet their daily driving needs. However, what if there is more than enough energy? Ford is looking closely at how battery electric vehicles can unlock additional value for customers in a world of a dynamic and intelligent electrical energy system.

This talk will briefly cover Ford's electrification strategy and the mission of Greenfield Labs, Ford's Silicon Valley outpost. It will review projects on connected charging infrastructure and hands-free charging and describe the ways that Ford researchers are envisioning smart vehicles in a smart world.

Date and Time:
Thursday, February 27, 2020 - 1:30pm
Venue:
Shriram 104

## Stanford's NetSeminar

Topic:
Abstract / Description:

Individuals of all genders invited to be a part of:
#StanfordToo: A Conversation about Sexual Harassment in Our Academic Spaces, where we will feature real stories of harassment at Stanford academic STEM in a conversation with Provost Drell, Dean Minor (SoM), and Dean Graham (SE3). We will have plenty of time for audience discussion on how we can take concrete action to dismantle this culture and actively work towards a more inclusive Stanford for everyone. While our emphasis is on STEM fields, we welcome and encourage participation from students, postdocs, staff, and faculty of all academic disciplines and backgrounds.

Date and Time:
Friday, April 19, 2019 - 3:30pm
Venue:
STLC 111

## John G. Linvill Distinguished Seminar on Electronic Systems Technology

Topic:
Internet of Things and Internet of Energy for Connecting at Any Time and Any Place
Abstract / Description:

In this presentation, I would like to discuss with you how to establish a sustainable and smart society through the internet of energy for connecting at any time and any place. I suspect that you have heard the phrase, "Internet of Energy" less often. The meaning of this phrase is simple. Because of a ubiquitous energy transmission system, you do not need to worry about a shortage of electric power. One of the most important items for establishing a sustainable society is [...]

"Inaugural Linvill Distinguished Seminar on Electronic Systems Technology," EE News, July 2018

Date and Time:
Monday, January 14, 2019 - 4:30pm
Venue:
Hewlett 200

## Claude E. Shannon's 100th Birthday

Topic:
Centennial year of the 'Father of the Information Age'
Abstract / Description:

From UCLA Shannon Centennial Celebration website:

Claude Shannon was an American mathematician, electrical engineer, and cryptographer known as "the father of information theory". Shannon founded information theory and is perhaps equally well known for founding both digital computer and digital circuit design theory. Shannon also laid the foundations of cryptography and did basic work on code breaking and secure telecommunications.

Events taking place around the world are listed at IEEE Information Theory Society.

Date and Time:
Saturday, April 30, 2016 - 12:00pm
Venue:
N/A

## NetSeminar

Topic:
BlindBox: Deep Packet Inspection over Encrypted Traffic
Abstract / Description:

SIGCOMM 2015, Joint work with: Justine Sherry, Chang Lan, and Sylvia Ratnasamy

Many network middleboxes perform deep packet inspection (DPI), a set of useful tasks which examine packet payloads. These tasks include intrusion detection (IDS), exfiltration detection, and parental filtering. However, a long-standing issue is that once packets are sent over HTTPS, middleboxes can no longer accomplish their tasks because the payloads are encrypted. Hence, one is faced with the choice of only one of two desirable properties: the functionality of middleboxes and the privacy of encryption.

We propose BlindBox, the first system that simultaneously provides both of these properties. The approach of BlindBox is to perform the deep-packet inspection directly on the encrypted traffic. BlindBox realizes this approach through a new protocol and new encryption schemes. We demonstrate that BlindBox enables applications such as IDS, exfiltration detection and parental filtering, and supports real rulesets from both open-source and industrial DPI systems. We implemented BlindBox and showed that it is practical for settings with long-lived HTTPS connections. Moreover, its core encryption scheme is 3-6 orders of magnitude faster than existing relevant cryptographic schemes.

Date and Time:
Wednesday, November 11, 2015 - 12:15pm to 1:30pm
Venue:
Packard 202

## NetSeminar

Topic:
Precise localization and high throughput backscatter using WiFi signals
Abstract / Description:

Indoor localization holds great promise to enable applications like location-based advertising, indoor navigation, inventory monitoring and management. SpotFi is an accurate indoor localization system that can be deployed on commodity WiFi infrastructure. SpotFi only uses information that is already exposed by WiFi chips and does not require any hardware or firmware changes, yet achieves the same accuracy as state-of-the-art localization systems.

We then talk about BackFi, a novel communication system that enables high throughput, long range communication between very low power backscatter IoT sensors and WiFi APs using ambient WiFi transmissions as the excitation signal. We show via prototypes and experiments that it is possible to achieve communication rates of up to 5 Mbps at a range of 1 m and 1 Mbps at a range of 5 meters. Such performance is an order to three orders of magnitude better than the best known prior WiFi backscatter system.

Date and Time:
Thursday, October 15, 2015 - 12:15pm to 1:30pm
Venue:
Gates 104

## NetSeminar

Topic:
TBA
Abstract / Description:

TBA

Date and Time:
Thursday, December 3, 2015 - 12:15pm to 1:30pm
Venue:
Gates 104

## NetSeminar

Topic:
BlindBox: Deep Packet Inspection over Encrypted Traffic
Abstract / Description:

SIGCOMM 2015, Joint work with: Justine Sherry, Chang Lan, and Sylvia Ratnasamy

Many network middleboxes perform deep packet inspection (DPI), a set of useful tasks which examine packet payloads. These tasks include intrusion detection (IDS), exfiltration detection, and parental filtering. However, a long-standing issue is that once packets are sent over HTTPS, middleboxes can no longer accomplish their tasks because the payloads are encrypted. Hence, one is faced with the choice of only one of two desirable properties: the functionality of middleboxes and the privacy of encryption.

We propose BlindBox, the first system that simultaneously provides both of these properties. The approach of BlindBox is to perform the deep-packet inspection directly on the encrypted traffic. BlindBox realizes this approach through a new protocol and new encryption schemes. We demonstrate that BlindBox enables applications such as IDS, exfiltration detection and parental filtering, and supports real rulesets from both open-source and industrial DPI systems. We implemented BlindBox and showed that it is practical for settings with long-lived HTTPS connections. Moreover, its core encryption scheme is 3-6 orders of magnitude faster than existing relevant cryptographic schemes.

Date and Time:
Wednesday, October 7, 2015 - 12:15pm to 1:30pm
Venue:
AllenX Auditorium

## Probability seminar presents "Gibbs measures and Glauber dynamics on trees"

Topic:
Gibbs measures and Glauber dynamics on trees
Abstract / Description:

The Ising model has two a priori different notions of equilibrium state: Gibbs states (which satisfy the DLR equations) and Glauber-invariant states (which are invariant under a natural dynamics). On the integer lattice these notions have long been known to be equivalent for shift-invariant states. Moreover, Holley showed in 1971 that any shift-invariant state converges weakly to the set of Gibbs states when evolved under Glauber dynamics. I will show how these results can be extended to infinite regular trees by defining a new notion of free energy density in the framework of sofic entropy theory.

Date and Time:
Monday, November 9, 2020 - 4:00pm
Venue:
Zoom ID 970 0451 1523 (+password)

## Workshop in Biostatistics presents "Algorithm-assisted decision making in child welfare"

Topic:
Algorithm-assisted decision making in child welfare
Abstract / Description:

● Counterfactual risk assessment, evaluation, and fairness
● Toward algorithmic accountability in public services
● Decisions in the presence of erroneous algorithmic scores
● Excerpt from Virginia Eubanks' Automating Inequality

Date and Time:
Thursday, November 12, 2020 - 2:30pm
Venue:
Zoom ID 926 9609 8893 (+password)

## Statistics Department Seminar presents "On the statistical foundations of adversarially robust learning"

Topic:
On the statistical foundations of adversarially robust learning
Abstract / Description:

Robustness has long been viewed as an important desired property of statistical methods. More recently, it has been recognized that complex prediction models such as deep neural nets can be highly vulnerable to adversarially chosen perturbations of their outputs at test time. This area, termed adversarial robustness, has garnered an extraordinary level of attention in the machine learning community over the last few years. However, little is known about the most basic statistical questions. In this talk, I will present answers to some of them. In particular, I will show how class imbalance has a crucial effect, and leads to unavoidable tradeoffs between robustness and accuracy, even in the limit of infinite data (i.e., for the Bayes error). I will also show other results, some of them involving novel applications of results from robust isoperimetry (Cianchi et al, 2011).

This is joint work with Hamed Hassani, David Hong, and Alex Robey.

Date and Time:
Tuesday, November 17, 2020 - 4:30pm

## Statistics Department Seminar presents "Conditional calibration for false discovery rate control under dependence"

Topic:
Conditional calibration for false discovery rate control under dependence
Abstract / Description:

We introduce a new class of methods for finite-sample false discovery rate (FDR) control in multiple testing problems with dependent test statistics where the dependence is fully or partially known. Our approach separately calibrates a data-dependent p-value rejection threshold for each hypothesis, relaxing or tightening the threshold as appropriate to target exact FDR control. In addition to our general framework we propose a concrete algorithm, the dependence-adjusted Benjamini–Hochberg (dBH) procedure, which adaptively thresholds the q-value for each hypothesis. Under positive regression dependence the dBH procedure uniformly dominates the standard BH procedure, and in general it uniformly dominates the Benjamini–Yekutieli (BY) procedure (also known as BH with log correction). Simulations and real-data examples illustrate power gains over competing approaches to FDR control under dependence.

This is joint work with Lihua Lei.

Joint Colloquium with Berkeley at Stanford

Date and Time:
Tuesday, November 10, 2020 - 4:30pm

## Statistics Department Seminar presents "Some theoretical results on model-based reinforcement learning"

Topic:
Some theoretical results on model-based reinforcement learning
Abstract / Description:

We discuss some recent results on model-based methods for reinforcement learning (RL) in both online and offline problems.

For the online RL problem, we discuss several model-based RL methods that adaptively explore an unknown environment and learn to act with provable regret bounds. In particular, we focus on finite-horizon episodic RL where the unknown transition law belongs to a generic family of models. We propose a model based "value-targeted regression" RL algorithm that is based on an optimism principle: in each episode, the set of models that are "consistent" with the data collected is constructed. The criterion of consistency is based on the total squared error that the model incurs on the task of predicting values as determined by the last value estimate along the transitions. The next value function is then chosen by solving the optimistic planning problem with the constructed set of models. We derive a bound on the regret, for an arbitrary family of transition models, using the notion of the so-called Eluder dimension proposed by Russo and Van Roy (2014).

Next we discuss batch-data (offline) reinforcement learning, where the goal is to predict the value of a new policy using data generated by some behavior policy (which may be unknown). We show that the fitted Q-iteration method with linear function approximation is equivalent to a model-based plugin estimator. We establish that this model-based estimator is minimax optimal and its statistical limit is determined by a form of restricted chi-square divergence between the two policies.

Date and Time:
Tuesday, November 3, 2020 - 4:30pm
Venue:
Registration required

## Statistics Department Seminar presents "Screenomics: A playground for the mining and modeling of both "big" and "small" longitudinal data"

Topic:
Screenomics: A playground for the mining and modeling of both "big" and "small" longitudinal data
Abstract / Description:

We recently developed and forwarded a framework for capturing, visualizing, and analyzing the unique record of an individual's everyday digital experiences: screenomics. In our quest to derive knowledge from and understand screenomes – ordered sequences of hundreds of thousands of smartphone and laptop screenshots obtained every five seconds for between one day and six months – the data have become a playground for learning about the computational machinery used to process images and text, machine learning algorithms, human-labeling of unknown taxonomies, qualitative inquiry, and the tension between N = 1 and N = many approaches. Using illustrative problems, I share how engagement with these new data is reshaping both how we do analyses and how we study the person-context transactions that drive human behavior.

Date and Time:
Tuesday, October 27, 2020 - 4:30pm

## Statistics Department Seminar presents "Two mathematical lessons of deep learning"

Topic:
Two mathematical lessons of deep learning
Abstract / Description:

Recent empirical successes of deep learning have exposed significant gaps in our fundamental understanding of learning and optimization mechanisms. Modern best practices for model selection are in direct contradiction to the methodologies suggested by classical analyses. Similarly, the efficiency of SGD-based local methods used in training modern models appeared at odds with the standard intuitions on optimization.

First, I will present the evidence, empirical and mathematical, that necessitates revisiting classical notions such as over-fitting. I will continue to discuss the emerging understanding of generalization and, in particular, the "double descent" risk curve, which extends the classical U-shaped generalization curve beyond the point of interpolation.

Second, I will discuss why the landscapes of over-parameterized neural networks are essentially never convex, even locally. Yet, they satisfy the local Polyak–Lojasiewicz condition, which allows SGD-type methods to converge to a global minimum.

A key piece of the puzzle remains: how do these lessons come together to form a complete mathematical picture of modern DL?

Date and Time:
Tuesday, October 20, 2020 - 4:30pm

## Statistics Department Seminar presents "Backfitting for large-scale crossed random effects regressions"

Topic:
Backfitting for large-scale crossed random effects regressions
Abstract / Description:

Large-scale genomic and electronic commerce data sets often have a crossed random effects structure, arising from genotypes x environments or customers x products. Naive methods of handling such data will produce inferences that do not generalize. Regression models that properly account for crossed random effects can be very expensive to compute. The cost of both generalized least squares and Gibbs sampling can easily grow as N^(3/2) (or worse) for N observations. Papaspiliopoulos, Roberts and Zanella (2020) present a collapsed Gibbs sampler that costs O(N), but under an extremely stringent sampling model. We propose a backfitting algorithm to compute a generalized least squares estimate and prove that it costs O(N) under greatly relaxed though still strict sampling assumptions. Empirically, the backfitting algorithm costs O(N) under further relaxed assumptions. We illustrate the new algorithm on a ratings data set from Stitch Fix.

This is based on joint with Swarnadip Ghosh and Trevor Hastie of Stanford University.

Date and Time:
Tuesday, October 13, 2020 - 4:30pm

## Statistics Department Seminar presents "Berry–Esseen bounds for Chernoff-type nonstandard asymptotics in isotonic regression"

Topic:
Berry–Esseen bounds for Chernoff-type nonstandard asymptotics in isotonic regression
Abstract / Description:

A Chernoff-type distribution is a non-normal distribution defined by the slope at zero of the greatest convex minorant of a two-sided Brownian motion with a polynomial drift. While a Chernoff-type distribution appears as the distributional limit in many nonregular estimation problems, the accuracy of Chernoff-type approximations has been largely unknown. In this talk, I will discuss Berry–Esseen bounds for Chernoff-type limit distributions in the canonical nonregular statistical estimation problem of isotonic (or monotone) regression. The derived Berry–Esseen bounds match those of the oracle local average estimator with optimal bandwidth in each scenario of possibly different Chernoff-type asymptotics, up to multiplicative logarithmic factors. Our method of proof differs from standard techniques on Berry–Esseen bounds, and relies on new localization techniques in isotonic regression and an anti-concentration inequality for the supremum of a Brownian motion with a Lipschitz drift.

This talk is based on joint work with Qiyang Han.

Date and Time:
Tuesday, October 6, 2020 - 4:30pm

## Probability Seminar presents "Non-stationary fluctuations for some non-integrable models"

Topic:
Non-stationary fluctuations for some non-integrable models
Abstract / Description:

The Kardar-Parisi-Zhang (KPZ) equation is a conjecturally universal model for dynamics of fluctuating interfaces such as fires and epidemic fronts. The universality was originally justified by Kardar, Parisi, and Zhang via non-rigorous renormalization group calculations. In this talk, we introduce some mathematically rigorous results and take a step towards this universality in the context of some non-integrable interacting particle systems outside their respective invariant measures.

Date and Time:
Monday, November 16, 2020 - 4:00pm

## SystemX Bonus Seminar: Apple Tech Talk: Mobile Silicon as the New Driver of Mainstream Computing

Topic:
Apple Tech Talk: Mobile Silicon as the New Driver of Mainstream Computing
Abstract / Description:

Mainstream computing has gone through multiple epochs, and the latest is now upon us. The importance of energy and space efficiency, power density, and the integration of application-driven acceleration has made mobile computing a new driver of compute architecture—and the next generation of mainstream computing.

Date and Time:
Tuesday, November 17, 2020 - 4:30pm
Venue:
Zoom

## SystemX presents "How AI can achieve deep understanding of text and speech: Structure-exploiting strategies for knowledge extraction"

Topic:
How AI can achieve deep understanding of text and speech: Structure-exploiting strategies for knowledge extraction
Abstract / Description:

Humans are good at engaging in conversation and understanding written text, so much so that these sorts of natural language processing (NLP) tasks are widely considered essential requirements for any truly intelligent artificial agent. The NLP community has made a great deal of progress over the last few years on certain challenges (for example, the SuperGLUE benchmarks and Winograd Schemas), thanks to the power of neural networks and pre-trained models like BERT and GPT-3. Nevertheless, existing systems still don't approach human performance for useful tasks such as answering general, open-ended questions about stories or participating in dialogue.

In this talk, I analyze why NLP's recent achievements do not generalize to true language understanding. I suggest a broad remedy – automated large-scale extraction of knowledge from text and reasoning with that knowledge, together with existing background information, to achieve deeper understanding – and propose methods to achieve this. I focus on texts that have exploitable structure, including bulleted lists and tables, and discuss the techniques I've used for automating understanding of regulatory and molecular biology texts. Finally, I discuss how comprehending the goal-action-effect structure implicit in narrative texts, including news articles and business case studies, can help extract useful general knowledge about actions and plans, and consider how these techniques could contribute to AI succeeding at the difficult tasks of story understanding and analysis.

Date and Time:
Thursday, November 19, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX presents "Advances on microwave and mmWave imaging"

Topic:
Advances on microwave and mmWave imaging
Abstract / Description:

Microwave and millimeter-wave (mmWave) Imaging Systems have witnessed huge progress during the last two decades, thanks to the advances in the semiconductor miniaturization in analog as well as digital circuits. Notwithstanding the ever-increasing demands for mmWave imaging solutions serving in the industrial, security, and medical domains, many efforts have been successfully accomplished to mature mmWave imaging technologies in order to qualify for the real world problems. Many more methods are still yet struggling to see the light away from the lab bench and the computer simulators. This talk aims to present and overview on latest mmWave multistatic imaging technologies and to help raise the awareness on the latest trends and challenges.

Date and Time:
Thursday, November 5, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX presents "The Investment Landscape for Quantum Computing"

Topic:
The Investment Landscape for Quantum Computing
Abstract / Description:

As we map out the frontiers of innovation, quantum computing (QC) looms large as possibly the most important scientific invention of the 21st century. QC upends what we thought to be fundamental principles of computer science, promising to execute entirely new classes of algorithms beyond the practical capabilities of conventional computers. The technology was initially explored in the 1980s when, at a series of lectures at MIT and Caltech, Berkeley Professor Paul Benioff defined the concept of a quantum circuit, and Richard Feynman floated the basic model for building such a computer. The technology had struggled to mature throughout the last few decades, but has started accelerating the recent years. In this talk, we will explore the current landscape of commercial entities pursuing the development of quantum computers and ponder on how they might change the future.

Date and Time:
Thursday, October 29, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX BONUS seminar presents "Fault-Tolerant Quantum Computing with Silicon Photonics"

Topic:
Fault-Tolerant Quantum Computing with Silicon Photonics
Abstract / Description:

Quantum computers promise a new paradigm of computation where information is processed in a way that has no classical analogue. However, the known problems for which quantum computers offer a computational advantage, require long gate sequences and large number of qubits. Error-correction codes and fault-tolerant gate implementation require the encoding of logical qubits on a large number of physical qubits, and given the overheads, it is expected that general purpose quantum computers will have millions of physical qubits, thus requiring an underlying qubit technology that can be manufactured at scale. Photons make great qubits, they are cheap to produce, resilient to noise and the only known option for quantum networks. Most crucially, they can be efficiently manipulated with silicon photonics, an intrinsically scalable and manufacturable platform in which all the fundamental quantum gates can be implemented. In this talk, I will describe an architecture for universal fault-tolerant quantum computing based on linear optics, in the process I will explain how measurement-induced non-linearity can overcome the challenge of creating entanglement and how loss can be effectively tackled with error correcting codes.

Date and Time:
Tuesday, October 27, 2020 - 2:00pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX presents "Interactive Artificial Intelligence — The Ultimate Challenge in Robotics"

Topic:
Interactive Artificial Intelligence — The Ultimate Challenge in Robotics
Abstract / Description:

In the last decade, we have experienced a significant leap forward in computer vision for tasks such as object recognition, reconstruction, 3D vision, detection, and tracking, where artificial systems have reached human and sometimes even "superhuman" performance. These advances result from the combination of novel machine learning algorithms, more powerful computers and large-enough curated datasets, which allow for off-line learning. This success has not yet transferred to robotics. In robotics, despite significant progress, machines are still far behind humans when it comes to physical and purposeful interaction with unstructured environments. Humans largely exploit physical interactions with the surroundings to solve complex long-horizon tasks (cooking, cleaning, tidying, assembling...) in "perceptually dirty", cluttered, uncontrolled, and often unknown environments such as homes, and offices. On the contrary, most of the state-of-the-art robotic solutions are restricted to short-horizon tasks in "perceptually clean" environments and try to minimize the interaction with the surroundings.

Despite all the challenges of physical interaction, in my research at SVL I advocate to consider interactions with the environment as part of the solution instead of the problem. In my talk I will present work in our group exploiting physical interaction to solve robotic tasks. I will present our work on Interactive Navigation, tasks where the robotic agent needs to interact with the environment (e.g. open doors, push away obstacles) to achieve a desired location. Solving this type of navigation is necessary to move in common human uncontrolled environments such as our homes. I will also present our work on Mechanical Search, where we equip robots with skills to search efficiently for target objects in piles of cluttered objects using physical interaction (pushing other objects, grasping them), a problem that is faced frequently not only in home robotics but also in logistic domains. And finally, I will present iGibson, SVL's large effort to provide interactive agents with a simulation environment to train and test interactive AI solutions, and demonstrate that our approach outperforms state-of-the-art learning-based and classical methods on real-world data while maintaining efficiency.

Date and Time:
Thursday, October 22, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX presents "Food Making Robots"

Topic:
Food Making Robots
Abstract / Description:

Over the next few years, food making robots are expected to be prevalent in our lives. The speaker describes this trend, shows video demonstrations of some commercially available food making robots and indicates what we could expect in the future.

Date and Time:
Thursday, October 15, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX presents "Quantum Computing – Where does the Hope meet the Hype?"

Topic:
Quantum Computing – Where does the Hope meet the Hype?
Abstract / Description:

Quantum computers exploit the bizarre features of quantum physics -- uncertainty, entanglement, and measurement -- to perform tasks that are impossible using conventional means, such as computing over ungodly amounts of data, and communicating via teleportation. I will describe the architecture of a quantum computer based on individual atomic clock qubits, suspended and isolated with electric fields, perfectly replicable with no idle errors, and individually addressed with laser beams. This leading physical representation of a quantum computer has allowed unmatched demonstrations of small algorithms and emulations of hard quantum problems with more than 50 quantum bits. While this system can solve some esoteric tasks that cannot be accomplished in conventional devices, it remains a great engineering challenge to build a quantum computer big enough to be generally useful for society. But the good news is that this is not a scientific challenge, as we know the technology needed and it's not quantum.

Date and Time:
Thursday, October 8, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX: Intelligent Sensors in Every Stream: Engineering Deployable Water Quality Sensors

Topic:
Intelligent Sensors in Every Stream: Engineering Deployable Water Quality Sensors
Abstract / Description:

Water contamination can harm human health and alter aquatic ecosystems. Current treatment only addresses collected wastewater, although most pollution occurs from uncollected water (e.g., agricultural runoff). Water quality is typically monitored at treatment plants with time-intensive, costly, analytical laboratory techniques that limit monitoring frequency and spatiotemporal resolution; exhibit dangerous lag times between sampling and measurement; and constrain monitoring to treatment plants. Presently, real-time water monitoring is restricted to indirect measures of pollution (e.g., pH, dissolved oxygen); we design modular sensors that directly measure pollutants such as ammonium, which contributes to harmful algal blooms. Our first step toward that vision is designing selective, durable, deployable ammonia sensors that achieve laboratory method detection (1 mg/L) at field sensor prices (<\$30) and breakthrough selectivity (>99%).

Date and Time:
Thursday, October 1, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode

## SystemX presents "Enterprise Software to Accelerate the Quantum Revolution"

Topic:
Enterprise Software to Accelerate the Quantum Revolution
Abstract / Description:

Through the transformative impact of quantum computing, powerful software and algorithms for the next generation of high-performance computing are being developed. These quantum algorithms will provide solutions to some of the world's toughest computational problems, including simulation, optimization, and machine learning applications in chemistry, finance, logistics, pharmaceuticals, engineering, and materials - in short, solve massive, real-world problems that will revolutionize industry and society.

Especially in recent years, we have witnessed a dramatic increase in the amount of investment and progress made in both quantum computing hardware and software architectures. This talk will look at the dramatic developments of the last decade and discuss the various approaches to the application of quantum computing. The goal is to suggest trends of how commercial quantum computing technology will evolve in the coming years and describe how this tremendous processing power could unleash exponential advances for humanity.

Date and Time:
Thursday, September 24, 2020 - 4:30pm
Venue:
Zoom ID: 920 2334 9868; +passcode