Evgenii Zheltonozhskii
Computer Scientist at Technion
Haifa District, Israel
Summary
🤩
Rockstar🎓
Top SchoolEvgenii Zheltonozhskii is a computer scientist and PhD candidate in condensed matter physics at Technion with 12 years of software engineering experience spanning research internships, freelance engineering, and open-source contributions. He holds an MSc in computer science with a focus on deep learning and self-supervised methods, and applies rigorous scientific thinking to practical ML and C++ library engineering. Evgenii has strengthened testing and CI for projects like tiny-dnn, modernized GAN code for TensorFlow compatibility, and improved core correctness in the xtensor C++ tensor library. His contributions to Google’s BIG-bench include enhancing the ARC task and evaluation metrics, showing an interest in benchmarking and model evaluation at scale. Comfortable both in low-level C++ and ML pipelines, he combines academic depth with hands-on bug-fixing and test automation skills. Based in Haifa, he brings a rare mix of condensed-matter research perspective and production-focused software craftsmanship.
12 years of coding experience
7 years of employment as a software developer
Bachelor’s Degree, Computer Science, Physics and Math, Bachelor’s Degree, Computer Science, Physics and Math at Technion - Israel Institute of TechnologyHavat HaNoar HaTsioni
Russian, English, Hebrew, German
Stackoverflow
898reputation
79kreached
42answers
4questions
Github Skills (25)
c-language10
python10
multidimensional-arrays10
testing10
machine-learning10
gt10
generative-adversarial-network10
numpy10
deep-learning10
tensorflow10
neural-network10
nlp10
tensor10
cprogramming-language10
metric10
Programming languages (18)
Github contributions (5)
tiny-dnn/tiny-dnn
Oct 2016 - Aug 2017
header only, dependency-free deep learning framework in C++14
Role in this project:
QA Engineer / Test Automation Engineer
QA Engineer / Test Automation EngineerContributions:55 commits, 80 PRs, 39 pushes in 10 months
Contributions summary:Evgenii's contributions primarily focused on enhancing the testing infrastructure and test coverage for the tiny-dnn deep learning framework. This includes the addition of GPU tests, the replacement of pico tests with gtests, and the implementation of tests utilizing Google Test (gtest) framework. The user also addressed and fixed various bugs within the test implementations to ensure the accuracy of the framework's functionalities.
cppheaderdeep-learningc-plus-plusmachine-learning
google/BIG-bench
May 2021 - May 2022
Beyond the Imitation Game collaborative benchmark for measuring and extrapolating the capabilities of language models
Role in this project:ML Engineer
ML EngineerContributions:2 reviews, 41 commits, 19 PRs in 1 year
Contributions summary:Evgenii primarily contributed to the development and improvement of the Abstraction and Reasoning Corpus (ARC) benchmark task within the Google BIG-bench repository. Their work included adding the ARC task, incorporating an edit-distance-based metric, and fixing issues related to example count calculations and task descriptions. Further contributions involved adding the 1-shot setting to the task and making the scaling plots runnable in colab, which suggests model evaluation and analysis. These changes indicate a focus on evaluating and refining the ARC task's functionality and evaluation capabilities within the BIG-bench framework.
bertmachine-learningbenchmarkmeasuringbenchmarks
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.