Badr Idrissi
Research Scientist at Meta
Greater Paris Metropolitan Region France
Summary
🤩
Rockstar🎓
Top SchoolBadr Idrissi is a research scientist with a decade of experience at the intersection of deep learning, NLP, interpretability and adversarial robustness, currently at Meta after a CIFRE PhD. He has a strong academic foundation from CentraleSupélec and ENS Paris-Saclay and a track record of impactful internships and research projects across FAIR, Inria, NAVER Labs Europe and more. His open-source contributions include improving image parameterizations and testing in the widely used lucid toolkit for neural network interpretability, highlighting a practical focus on reproducible visualization tools. Badr combines rigorous mathematical training with hands-on experimentation—ranging from audio feature visualization to adversarial training in machine translation—and has translated research into patents and publications. He’s particularly adept at bridging theory and engineering to make model behavior more interpretable and robust, often producing tooling and baselines that others can build on.
10 years of coding experience
1 year of employment as a software developer
Gap Year, Gap Year at Digital Tech YearMaster of Science - MS, Machine Learning, Applied Mathematics, Master of Science - MS, Machine Learning, Applied Mathematics at École normale supérieure Paris-SaclayMaster of Engineering - MEng, Mathematics and Computer Science, Master of Engineering - MEng, Mathematics and Computer Science at CentraleSupélecBachelor of Science - BSc, Bachelor of Science - BSc at Lycée Pierre Corneille
French, Arabic, English
Stackoverflow
126reputation
376reached
1answer
0questions
Github Skills (9)
machine-learning10
visualization10
interpretation10
tensorflow10
visualizations10
python9
jupyter-notebook8
testing8
tkinter6
Programming languages (4)
Github contributions (5)
tensorflow/lucid
Jun 2019 - Jun 2019
A collection of infrastructure and tools for research in neural network interpretability.
Role in this project:
ML Engineer
ML EngineerContributions:11 commits, 1 PR, 8 comments in 21 days
Contributions summary:Badr primarily contributed to the development and improvement of image parameterizations within the Lucid framework, a toolset for neural network interpretability. They addressed a bug related to file encoding in a test. Their work included fixing image width/height switching, updating the image parameterization to accept an arbitrary number of channels, and adding integration tests, which demonstrates a focus on improving visualization capabilities and testing the changes. These modifications directly support the project's goal of enhancing the interpretability of machine learning models.
pytorchinterpretabilitydeep-learningjupyter-notebookinfrastructure
BadrYoubiIdrissi/TIPE-Algorithme-Genetique
May 2016 - Feb 2017
Contributions:8 PRs, 39 pushes, 13 branches in 9 months
roboticspythonrobotneatneat-algorithm
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.