Bio

I am a third-year PhD student in computer science at MIT, where I research topics in artificial intelligence, applied cryptography, and technology policy. My current research focus, the lottery ticket hypothesis, involves exploring the basic properties of neural networks by developing techniques that make it possible to find small neural networks that train successfully. I also design and teach a programming course for law students at the Georgetown University Law Center

Experience
2018-19: Researcher Sparsity for NLP Transfer Learning Google Brain
2019: Professor Adjunct Professor of Law, Lead Instructor (Programming for Lawyers) [class] Georgetown University Law Center
2018: Intern Program Synthesis Google Brain
2018: TA Teaching Assistant (Computer and Network Security) [class] MIT
2017: Intern Fully Homomorphic Encryption [code] Google
2016: Professor Adjunct Professor of Law, Co-Instructor (Programming for Lawyers) [class] Georgetown University Law Center
2015-16: Fellow Staff Technologist, Center on Privacy and Technology Georgetown University Law Center
2015: TA Teaching Assistant (General Computer Science - COS126) [class] Princeton University
2014: TA Teaching Assistant (Information Security - COS432) [class] Princeton University
2014: Intern Encryption Key Management Infrastructure Google
2013: Intern Project Siena (Pre-Beta) Compiler Team Microsoft
2012: Intern Automated Test and Retest Innovative Defense Technologies
Papers, Press, & Publications
2019
  • The Lottery Ticket Hypothesis at Scale (Jonathan Frankle, Gintare Karolina Dziugaite, Daniel M. Roy, and Michael Carbin) [arxiv]
  • ICLR Best Paper Award: The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks (Jonathan Frankle and Michael Carbin) [arxiv | openreview]
  • ICLR Workshop on Debugging Machine Learning Models: Dissecting Pruned Neural Networks (Jonathan Frankle and David Bau) [paper]
  • No Starch Press: Computer Programming for Lawyers (Jonathan Frankle and Paul Ohm) [website]
2018
  • Usenix Security: Practical Accountability of Secret Processes (Jonathan Frankle, Sunoo Park, Daniel Shaar, Shafi Goldwasser, Daniel Weitzner) [eprint | usenix ]
  • Florida Law Review: Desirable Inefficiency (Paul Ohm and Jonathan Frankle) [article]
2016
  • Interview with NPR: Researchers Find Flaws in Police Facial Recognition Technology [interview]
  • Report: The Perpetual Lineup: Unregulated Police Face Recognition in America (Clare Garvie, Alvaro Bedoya, Jonathan Frankle) [report]
  • Washington Post article about my class: Government Lawyers Don’t Understand the Internet [article]
  • The Atlantic: How Russia’s New Facial Recognition App Could End Online Anonymity (Jonathan Frankle) [article]
  • The Atlantic: Facial-Recognition Software Might Have a Racial Bias Problem (Clare Garvie and Jonathan Frankle) [article]
  • POPL: Example-Directed Synthesis: A Type-Theoretic Interpretation (Jonathan Frankle, Peter-Michael Osera, Dave Walker, Steve Zdancewic) [pdf | github | artifact | talk]
2015
  • MSE Thesis: Type-Directed Synthesis of Products [pdf | arxiv]
2014
  • BSE Thesis: Programming Recursive Software-Defined Networks [pdf]
  • Course Project: Why King George III Can Encrypt [pdf | blog post]
Education
2016-21: PhD Computer Science  |  Neural Networks MIT
2014-15: MSE Computer Science  |  Programming Languages Princeton University
2011-14: BSE Computer Science   Princeton University
Service
2018-19 Invited Expert: Expert Group on Artificial Intelligence OECD
2018-22 Computer Science Department Advisory Council Princeton University