Roland Zimmermann
Senior Research Scientist at Google DeepMind
Zurich, Zurich, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolRoland Zimmermann is a Senior Research Scientist with 11 years of experience bridging physics and machine learning, currently working on AGI safety and frontier models at Google DeepMind. He completed a PhD at the University of Tübingen/IMPRS-IS after industry stints at Volkswagen, BMW and two research internships at Google Brain, bringing a pragmatic, production-aware perspective to research. His work spans adversarial robustness, model compression and core ML algorithms, with notable open-source contributions such as implementing and stabilizing the GenAttack adversarial attack in the widely used Foolbox toolbox (including JAX support and thorough test coverage). Trained originally as a physicist at Göttingen, he combines strong theoretical rigor with hands-on engineering across PyTorch, TensorFlow and JAX. Based in Zurich, he favors reproducible, well-tested research that transfers into industry practice. An often-overlooked strength is his track record of improving attack stability and testability—skills that make his ML research unusually robust and deployable.
11 years of coding experience
1 year of employment as a software developer
Doctor of Philosophy - PhD Physics Machine Learning, Doctor of Philosophy - PhD Physics Machine Learning at University of TübingenMaster's degree Physics, Master's degree Physics at The University of Göttingen
Stackoverflow
4,921reputation
365kreached
118answers
37questions
python
top-5%
machine-learning
top-5%
neural-network
top-5%
keras
top-5%
Github Skills (14)
pytorch10
machine-learning10
adversarial-attacks10
jax10
python10
tensorflow9
neural-network9
keras9
legend6
deep-learning6
numpy6
matplotlib6
scikit-learn6
encoding6
Programming languages (11)
Github contributions (5)
bethgelab/foolbox
Mar 2020 - Apr 2022
A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX
Role in this project:
ML Engineer
ML EngineerContributions:3 releases, 14 reviews, 80 commits in 2 years 1 month
Contributions summary:Roland implemented the GenAttack adversarial attack, including its core logic, type annotations, and unit tests. They added support for JAX models, and refactored and improved the attack's stability and test coverage. The user's commits demonstrate a focus on developing and refining an adversarial example generation algorithm for machine learning models.
pytorchadversarial-attackspythondeep-learningadversarial
zimmerrol/foolbox
Feb 2020 - Mar 2024
Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, Keras, …
Contributions:275 pushes, 104 branches in 4 years 1 month
pytorchpythondeep-learningadversarialtoolbox
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.