Maksym Andriushchenko
Principal Investigator
Tübingen, Baden-Württemberg, Germany
Summary
🤩
Rockstar🎓
Top SchoolMaksym Andriushchenko is a research-driven AI safety and alignment leader with a decade of experience bridging cutting-edge machine learning research and reliable engineering. Currently a Principal Investigator at the ELLIS Institute Tübingen and an Independent Group Leader at the Max Planck Institute for Intelligent Systems, he leads efforts on aligning advanced models with robust, verifiable behavior. His PhD and postdoc work at EPFL, supported by Google and Open Phil fellowships, underpin a track record of rigorous experimentation and reproducible benchmarks. As an active open-source contributor, he helped extend RobustBench—NeurIPS-recognized adversarial robustness benchmarks—by integrating new models and improving the model zoo and testing utilities. Colleagues describe him as someone who pairs deep theoretical insight with practical tooling that makes robustness research usable by the broader community.
10 years of coding experience
1 year of employment as a software developer
Master's thesis, Machine Learning, Master's thesis, Machine Learning at University of TübingenDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at EPFLMaster's degree, Computer Science, Master's degree, Computer Science at Universität des Saarlandes
Github Skills (14)
robust10
pytorch10
adversarial-machine-learning10
machine-learning10
eval10
python10
evaluation10
data-science10
robustness10
preprocess9
preprocessing9
dataprep9
data-preprocessing9
data-loading9
Programming languages (4)
Github contributions (5)
RobustBench/robustbench
Jun 2020 - Aug 2022
RobustBench: a standardized adversarial robustness benchmark [NeurIPS 2021 Benchmarks and Datasets Track]
Role in this project:
ML Engineer & Data Scientist
ML Engineer & Data ScientistContributions:1 release, 8 reviews, 240 commits in 2 years 2 months
Contributions summary:Maksym primarily contributed to the integration and evaluation of machine learning models within the adversarial robustness benchmark. They added a new model ("carmon\_et\_al\_2019") with an automated download feature. Furthermore, the user updated the codebase to support a new model ("carmon2019unlabeled") and restructured the model zoo, creating a data loading utility. In addition, the user implemented and tested code for clean accuracy testing and added tests for a new fast test.
pytorchbenchmarkingrobustnessadversarial-machine-learningdeep-learning
max-andr/joint-cnn-mrf
Jan 2018 - Feb 2018
Contributions:28 commits, 30 pushes, 2 branches in 1 month
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.