October 2021 - talks - lists.cs.princeton.edu

8:30pm Monday October 11 panel on DoJ China Initiative (on zoom, registration required)
by Jennifer Rexford 09 Oct '21

09 Oct '21

Title: China Initiative Panel Discussion Date/time: 8:30pm Monday October 11 Location: on Zoom, registration required Link to register: https://ccc.princeton.edu/events/china-initiative-panel-discussion The U.S. Department of Justice’s China Initiative was created in November 2018 by the Trump administration to focus on the investigation and prosecution of trade secret theft and economic espionage stemming from the Chinese government. Instead of focusing on espionage, the China Initiative has been targeting, investigating and prosecuting scientists, without espionage evidence, for failures to disclose collaboration activities with Chinese scientists during the times the activities themselves were encouraged. Such investigations and its chilling effect have affected many scientists at many universities, including Princeton. Recently, 177 faculty members wrote an open letter to Attorney General Merrick B. Garland to terminate the DOJ’s China Initiative and develop an alternative response to address the competition from China. In order to understand the China Initiative and its impact, the center for contemporary China and School of Engineering and Applied Science are co-hosting this Zoom panel to discuss the China initiative and its impact. We have invited several distinguished guests including Prof. Steven A. Kivelson of Stanford University who is one of the people drafted their open letter, Prof. Yiguang Ju of Princeton MAE department to talk about the chilling effect of the China Initiative, Prof. Rory Truex of Princeton Politics and International Affairs who conducted extensive research on the China Initiative, and Peter R. Zeidenberg, a Partner of Arent Fox who is an experienced defense atoner who have represented many scientists being prosecuted under the China Initiative and have published articles about the China Initiative. The panel discussion will be moderated by Prof. Yu Xie who is the director of Paul and Marcia Wythes Center for Contemporary China.

1 0

Prakash Murali will present his FPO "Enabling Practical Quantum Computation: Compiler and Architecture Techniques for Bridging the Algorithms-to-Devices Resource Gap" on Friday, October 15, 2021 at 2PM via Zoom.
by jfarquer＠cs.princeton.edu 08 Oct '21

08 Oct '21

Prakash Murali will present his FPO "Enabling Practical Quantum Computation: Compiler and Architecture Techniques for Bridging the Algorithms-to-Devices Resource Gap" on Friday, October 15, 2021 at 2PM via Zoom. Zoom link: https://princeton.zoom.us/j/92941593235 <https://www.google.com/url?q=https://princeton.zoom.us/j/92941593235&sa=D&s ource=calendar&usd=2&usg=AOvVaw0-s50q8TMhlZ6oB6v73fzs> The members of his committee are as follows: Margaret Martonosi (Adviser, Reader, Examiner); Readers: Kyle Jamieson and Fred Chong (University of Chicago); Examiners: Andrew Houck (Princeton EE), Nathalie de Leon (Princeton EE) A copy of his thesis is available upon request. Please email gradinfo(a)cs.princeton.edu <mailto:gradinfo@cs.princeton.edu> if you would like a copy of the thesis. Everyone is invited to attend his talk. Abstract follows below: Quantum computing (QC) is poised to fundamentally change what is computable in several domains. From the first 1- and 2-qubit systems in the early 2000s, today's QC hardware landscape includes cloud-accessible systems with 10-50 qubits and multiple qubit technology candidates. In spite of rapid hardware progress, the first practically useful QC applications have not been demonstrated yet, even though hundreds of QC algorithms have been developed in the last three decades. This is fundamentally because of a large gap between the resource requirements of QC applications and the capabilities of quantum hardware that is buildable in the near-term; qubit counts and operational noise constraints of applications exceed hardware capabilities by 5-6 orders of magnitude. This dissertation seeks to close the resource gap between quantum algorithms and hardware by developing quantum architecture and compilation techniques. Unlike related research efforts which largely focus on designing individual layers in the QC execution stack in isolation, this dissertation develops a cross-cutting approach to guide the design of the QC stack. Using this approach, Part I of this dissertation develops noise-adaptive compilation and crosstalk mitigation techniques that offer one to two orders of magnitude improvement in application fidelity compared to vendor compilers. Part II presents an extensive cross-platform architectural study of real QC systems and applies the cross-cutting design approach to architect trapped ion systems and tackle fundamental issues in quantum instruction set design. Part II offers up to four orders of magnitude improvement in reliability for future QC devices. These contributions have already influenced industry toolflows and architectures, with many vendors adopting noise-adaptive compilation, and adjusting their architecture and benchmarking practices in line with this dissertation's recommendations. This dissertation shows that cross-cutting design offers several orders of magnitude improvement in fidelity for QC systems, compared to existing approaches. Going forward, the research contributions and directions laid out in this dissertation have the potential to accelerate the progress towards practically viable QC by several years, rather than relying solely on hardware or application improvements.

1 0

Invitation to the 10/18 Columbia Center of AI Tech Symposium
by Scott Karlin 06 Oct '21

06 Oct '21

Inaugural Columbia Center of AI Tech Symposium Monday, October 18 , 2021 11:00 PT/2:00 ET - 2:00 PT/5:00 ET Online The Columbia Center of AI Technology (CAIT) in collaboration with Amazon was founded in 2020 with a mission to better society through the development and adoption of advanced AI technologies contributing to a more secure, connected, creative, sustainable, healthy, and equitable humanity. A major goal within this mission is to create a community of scholars and practitioners at the leading edge of AI and AI technology development. The Inaugural Annual CAIT Symposium will bring together researchers, leaders, and experts from academia, industry, and government to discuss advances and new challenges in the field of AI technology. RSVP: [ https://www.eventbrite.com/e/cait-inaugural-symposium-tickets-170360727522?… | https://www.eventbrite.com/e/cait-inaugural-symposium-tickets-170360727522?… ] Visit the CAIT Symposium Webpage: [ https://2021caitsymposium.splashthat.com/ | https://2021caitsymposium.splashthat.com/ ] Schedule (ET) 2:00 – 2:05 pm || Opening Remarks 2:05 – 2:30 pm || Keynote: Spyros Matsoukas, Amazon Alexa AI 2:30 – 2:55 pm || Keynote: Angeliki Metallinou, Amazon Alexa AI 2:55 – 3:10 pm || Spark Talk: Elias Bareinboim, Columbia Engineering, Computer Science Department 3:10 – 3:25 pm || Spark Talk: Differential Privacy: Challenges and Opportunities | Rachel Cummings, Columbia Engineering, Industrial Engineering and Operations Research Department 3:25 – 3:40 pm || Spark Talk: AI and New Abilities: Video Games for Blind Players | Brian Smith, Columbia Engineering, Computer Science Department 3:40 – 3:55 pm || Spark Talk: Active Scene Understanding with Robot Interactions | Shuran Song, Columbia Engineering, Computer Science Department 3:55 – 4:00 pm || Break 4:00 – 5:00 pm || Fireside Chat: Multi-sector Partnerships for Advancing AI Shih-Fu Chang, Interim Dean, Columbia Engineering Parag R. Chitnis, Associate Director for Programs, Acting Associate Director for Operations, USDA – National Institute of Food and Agriculture Erwin Gianchandani, Senior Advisor for Translation, Innovation, and Partnerships, National Science Foundation Prem Natarajan, Vice President, Amazon Alexa AI

1 0

Barak Nehoran will present his General Exam "The Computational Power of Quantum Interaction Through Classical Channels" on Friday, October 8, 2021 at 1PM via Zoom.
by jfarquer＠cs.princeton.edu 04 Oct '21

04 Oct '21

Barak Nehoran will present his General Exam on Friday, October 8, 2021 at 1PM via Zoom. Zoom Link: <https://princeton.zoom.us/j/92536310724?pwd=TjI5VytxWlV0M1I1TlRmL2duNjB0dz0 9> https://princeton.zoom.us/j/92536310724?pwd=TjI5VytxWlV0M1I1TlRmL2duNjB0dz09 The members of his committee are as follows: Ran Raz(adviser), Mark Zhandry(adviser), and Gillat Kol. Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. Title: The Computational Power of Quantum Interaction Through Classical Channels Abstract: https://docs.google.com/document/d/1kpLoZzUbynghf1pYy-Hbph7KT6voWijX9FnrdUoj _V8/edit?usp=sharing Reading List: https://docs.google.com/document/d/1-uw4ESeKypRI2kJtH2lihD3nmP5r8KCGr2RTtwkE fdE/edit?usp=sharing

1 0

Jordan Holland will present his Pre-FPO "New Directions in Network Traffic Analysis" on Thursday, October 7, 2021 at 1PM via Zoom.
by jfarquer＠cs.princeton.edu 04 Oct '21

04 Oct '21

Jordan Holland will present his Pre-FPO " New Directions in Network Traffic Analysis" on Thursday, October 7, 2021 at 1PM via Zoom. Zoom link: https://princeton.zoom.us/j/91304375990 Committee: Examiners: Nick Feamster (Adviser), Prateek Mittal, Ravi Netravali Readers: Jonathan Mayer and Vitaly Shmatikov (Cornell) All are welcome to attend. Title: New Directions in Network Traffic Analysis Abstract: The field of network traffic analysis has long focused on creating bespoke solutions for specific problems. More recently, the field has examined the effectiveness of machine learning techniques for a wide range of traffic analysis problems. This approach follows a specific pattern: collect network traffic, create a data processing pipeline to appropriately separate the traffic (e.g. traffic flows, applications, devices) and attach labels to the separated traffic, hand-engineer features specific to the task, and finally train models on the engineered features. This methodology, combined with the number and variety of problems comprising network traffic analysis, has led to the creation of custom analysis pipelines for each task, requiring custom data processing pipelines, new features, and new models. In this talk, we discuss the feasibility of using generalizable techniques for multiple steps in the typical traffic analysis pipeline. First, we introduce nPrint, a tool that generates a unified packet representation that is amenable for representation learning and model training. We integrate nPrint with automated machine learning (AutoML), resulting in nPrintML, an open-source system that eliminates manual feature engineering and model selection for a number of problems. Next, we examine how custom data processing pipelines affect reproducibility in the field. We inspect work leveraging multiple popular datasets and find that current practices render both comparing and reproducing work difficult. We then introduce pcapML, an open-source system that standardizes traffic analysis tasks at the dataset level, improving reproducibility and lowering the barrier to entry for testing new traffic analysis techniques.

1 0