Lev Mckinney
Old Toronto, Ontario, Canada
Summary
🤩
Rockstar🎓
Top SchoolLev Mckinney is a machine learning researcher and software engineer with nine years of hands-on experience, currently pursuing advanced graduate studies in Computer Science at the University of Toronto. He has contributed to model-based RL and imitation learning research—implementing and testing reward-learning innovations in notable open-source PyTorch projects—and has interned at PayPal building content personalization and NLG tooling. His research roles at CHAI and FAR AI reflect a focus on safe, practical ML and AI safety, and he is co-authoring work on world models that plan without reconstruction loss. Aviation-trained and futurist-minded, Lev blends rigorous academic performance (3.91 GPA) with production-focused engineering and a knack for refactoring complex ML systems into testable, extensible code.
9 years of coding experience
2 years of employment as a software developer
Comox Cadet Flying Training CentreDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Toronto
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (20)
imitation-learning10
pytorch10
serializable10
serializer10
python10
machine-learning10
serialization10
deserialization10
testing9
operation9
pytest9
numpy9
tensorrt9
tensorflow9
tensor9
Programming languages (6)
Github contributions (5)
HumanCompatibleAI/imitation
Jun 2022 - Sep 2022
Clean PyTorch implementations of imitation and reward learning algorithms
Role in this project:
ML Engineer
ML EngineerContributions:123 reviews, 85 commits, 18 PRs in 3 months
Contributions summary:Lev primarily contributed to the development and improvement of imitation and reward learning algorithms implemented in PyTorch. Their work involved significant refactoring and testing of reward functions, including the implementation of an EMA normalization layer. They also added reward ensembles and conservative reward functions, along with related trainers and loss functions for preference comparison algorithms. The user's contributions included integrating new features, addressing bugs, and improving the testing infrastructure.
pytorchimplementationsreinforcement-learningcleanmachine-learning
AlignmentResearch/tuned-lens
Feb 2023 - Jan 2025
Tools for understanding how transformer predictions are built layer-by-layer
Contributions:46 reviews, 125 PRs, 355 pushes in 1 year 10 months
pytorchnlpunderstandingdeep-learningmachine-learning
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.