Xavier Bouthillier
Research Scientist at Mila - Quebec Artificial Intelligence Institute
Montreal, Quebec, Canada
Summary
🤩
Rockstar🎓
Top SchoolXavier Bouthillier is a research scientist at Mila with 13 years of experience bridging deep learning research and production-grade tooling. A PhD student under Pascal Vincent, he has worked across NLP, computer vision, optimization and theoretical deep learning while focusing now on improving research methodology and reproducibility. He leads Oríon, a hyper-parameter optimization platform, and develops Mahler to automate large-scale experiments and statistical validation. A long-time contributor to foundational ML tooling, he helped advance Theano and Pylearn2—work that underpins much of modern Python-based deep learning infrastructure. Based in Montreal, he combines rigorous academic research with hands-on engineering and mentorship of students contributing to open-source projects.
13 years of coding experience
2 years of employment as a software developer
Bachelor's degree, Computer Science, Bachelor's degree, Computer Science at Université de MontréalExchange student, Computer science, Exchange student, Computer science at Albert-Ludwigs-Universität Freiburg im Breisgau
French, English, German
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (7)
randomization10
machine-learning10
unit-test10
python10
theano10
numpy10
deep-learning9
Programming languages (6)
Github contributions (5)
lisa-lab/pylearn2
May 2013 - Jun 2015
Warning: This project does not have any current developer. See bellow.
Role in this project:
ML Engineer
ML EngineerContributions:150 commits, 5 PRs, 5 pushes in 2 years
Contributions summary:Xavier's commits primarily focus on adding and modifying features related to training and monitoring machine learning models within the PyLearn2 framework. The user added a new channel for monitoring the time per epoch during the training process. The commits also include changes to the cost functions, such as the addition of SML (PCD) and CD-k methods for training RBMs. Furthermore, the user made modifications related to the probabilistic max pooling.
javascripttypescript
Theano/Theano
Feb 2014 - May 2017
Theano was a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. It is being continued as PyTensor: www.github.com/pymc-devs/pytensor
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:26 commits, 24 PRs, 15 pushes in 3 years 3 months
Contributions summary:Xavier primarily contributed to the Theano library by adding functionality for random number generation, specifically focusing on the Poisson distribution. This involved implementing helper functions and wrappers for NumPy's poisson, integrating it into Theano's random number streams, and creating unit tests to ensure correctness. Furthermore, the user addressed negative axis handling in the `Join` and `GpuJoin` operations, adding tests for variable and constant axis, which improved the functionality. The commits include optimizations and relative tests, suggesting an effort to improve performance and reliability of the library.
python-librarymathmulti-dimensionalpythonevaluate
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.