Who am I

I'm a postdoc at Facebook AI Research in Menlo Park, CA.
I investigate the learning dynamics of neural networks, and bring my scientific approach to Machine Learning research.
I was previously a Physics PhD student at Yale University.
I studied experimental High Energy Physics as a member of the ATLAS experiment at CERN.
In my spare time, I teach myself new foreign languages and data visualization techniques.
I truly enjoy learning.
I believe in reproducible research.
I love sports, technology, fine dining, and design.


my skills

Computing

I teach machines using: ,
PyTorch, Keras, TensorFlow, Theano.
I sometimes program in: , ,
Matlab, LabView, R, ROOT.
I have experience using: , , .

Math

Thorough knowledge of Group Theory,
Linear Algebra, Complex Analysis,
Statistics, and Data Analysis
from academic coursework and
research experience.

People Skills

Management, Event Planning, Effective
Communication, Active Listening,
Leadership, and Flexibility.


Would you like to know more about my skills and experience?

My Resume My CV
FAIR

Facebook AI Research (FAIR)

I am currently a Postdoctoral Researcher at Facebook Artificial Intelligence Research in Menlo Park, CA. My work focuses on understanding the dynamics of learning in neural networks, with a particular interest towards initialization techniques, generalization, and characterization of loss landscapes. I am also interested in fairness, interpretability, and uncertainty estimation in machine learning. If you share my research interests and you would like to find out more about FAIR internship opportunities, contact me.

LBNL view

Berkeley National Lab

When I'm not at FAIR, you can find me at LBNL where I'm currently a visiting Research Affiliate. I collaborate with colleagues from ATLAS and NERSC (the National Energy Research Scientific Computing center) on projects at the intersection of physics and computing. Our most recent joint work focuses on training Generative Adversarial Networks (GANs) to speed up High Energy Physics simulation.

ATLAS Detector CERN

ATLAS Experiment at CERN

I used to be located in Geneva, Switzerland where I conducted research for my PhD thesis as a member of the ATLAS collaboration at CERN (European Council for Nuclear Research). I developed software for the optimization of particle identification for the flavor tagging group, and I was involved in the LHC-wide Machine Learning working group to push for the adoption of modern ML in High Energy Physics and the improvement of our current libraries and tools. My analysis focused on the search of Higgs boson pairs decaying into $\gamma\gamma$bb.

Yale ATLAS Tipton Group

Yale University

At Yale, I worked with Prof. Tobias Golling on the design and application of Deep Learning algorithms to develop taggers for the identification of top quarks and gauge bosons in boosted topologies. I continued on as a member of Prof. Paul Tipton's group, and I worked on a qualification task to optimize b-tagging techniques, by including track-jets for Run II at the LHC (Large Hadron Collider). At Yale, I was a Teaching Fellow for 2 years, a McDougal Fellow and a member of the Graduate Student Assembly.

Berkeley Perlmutter

UC Berkeley

I graduated from Cal in 2013 with a double major in Astrophysics and Physics. During my undergraduate studies, I explored different research fields, including theoretical, observational and computational galactic dynamics, search for extraterrestrial intelligence at the SETI Institute, space weather at the Space Sciences Lab, and antimatter production with the AEgIS experiment at CERN. At Berkeley, I discovered my passion for teaching while tutoring at the Athletic Study Center and designing a new course named Sense & Science & Sensibility with professors Saul Perlmutter, Rob MacCoun and John Campbell.

Interests

When I'm not hunting Higgs pairs, I entertain myself by participating in outreach events
and educating myself and others on topics in science, statistics and technology.

  • Congressional Liaison

    I participate as a US LUA delegate in the annual meeting with Congress in Washington, DC, to advocate for funding for high energy physics and present reports on recent achievements and discoveries.

  • Datasauri

    Datasauri is a group of young physicists, data scientists and software developers that came together during the 2015 THE Port Hackathon at CERN and Campus Biotech with the intent of solving the problem of organization of humanitarian data.

  • HackingEDU

    I participated in the world largest educational hackathon in San Mateo, CA. My team developed an app that uses the MyScript API as a basis for handwritten mathematical formula recognition from photos using your portable device.

  • TEDxCERN Volunteer

    TEDxCERN is one of the many exciting events organized at CERN every year. As an assistant stage manager, I helped the organizers during the event, got the chance to interact with absolutely inspiring speakers in the backstage, and assisted them during rehersals.

  • CERN Guide

    As an official CERN guide, I lead tours of tourists and students who visit the laboratory. I conduct visits in English and Italian. My goal is to leave visitors with a better understanding of the importance of fundamental research, and curiousity towards High Energy Physics.

  • Pop Science - Nuit des Chercheurs

    A scientifically educated public is an empowered and responsible public. During the Nuit des Chercheurs, CERN volunteers set up a day of events in a large mall with the objective of making physics more accessible and fun for shoppers and passersby.

  • CERN S'Cool Lab

    S'Cool lab gives me the opportunity to pursue my passion for innovative science education during my residency at CERN. Tutors such as myself welcome middle and high school students, as well as groups of teachers, and lead lab sections in which attendees will perform experiments with cloud chambers, X-rays and electron tubes.

  • Lifelong Learning

    MOOCs are an innovative way of making higher education available to millions of people around the world. While I personally use platforms like Coursera and Udacity to continue learning, I also plan to try to design my own MOOC in the future.

  • Virtual Reality

    My next goal is to start creating virtual realities for educational and commercial purposes, to enhance classroom experience and give people access to environments and equipments they wouldn't normally be able to benefit from. I also collaborated on a VR short movie about the painful experience of OCD in simple daily tasks.

Upcoming Dates

  • Mar 13 - 14

    National Academy of Sciences

  • Apr 25 - 26

    Microsoft Research

  • May 6 - 9

    New Orleans, LA

  • Jul 15 - 19

    Lawrence Berkeley National Lab

  • Contact me

    Feel free to contact me via email or follow me on social media using the links below. Reach out to me with teaching and speaking opportunities, deep learning projects, or to intern with me at Facebook AI Research.