George Tucker
Staff Research Scientist at Google
Mountain View, California, United States
Summary
👤
Senior🎓
Top SchoolGeorge Tucker is a Staff Research Scientist with 13 years of experience advancing reinforcement learning and variational methods at Google Brain, where he established the widely used D4RL offline RL benchmark and developed state-of-the-art offline RL algorithms. He blends deep theoretical contributions—published at NeurIPS, ICML, and ICLR—with pragmatic engineering, contributing to major open-source projects like DeepMind’s Acme and Google Research codebases to ship robust RL components. Earlier, he built end-to-end wake-word detection systems at Amazon Echo, owning tooling, modeling, and on-device QA to improve customer experience. He mentors junior researchers, leads cross-functional efforts, and is a go-to expert on variance reduction and discrete latent-variable methods. Based in Mountain View, he pairs an MIT Ph.D. in mathematics with a knack for turning complex probabilistic ideas into production-grade code.
13 years of coding experience
11 years of employment as a software developer
Ph.D., Mathematics, Ph.D., Mathematics at Massachusetts Institute of TechnologyBachelor of Science (B.S.), Mathematics, Computer Science, Bachelor of Science (B.S.), Mathematics, Computer Science at Harvey Mudd College
Stackoverflow
33reputation
3kreached
0answers
2questions
Github Skills (28)
data-analysis10
convolutional-neural-networks10
python10
model-driven10
machine-learning10
reinforcement-learning10
model-building10
agent10
deep-learning10
tensorflow10
neural-network10
modeling10
model-driven-development10
data-structure9
algorithm9
Programming languages (2)
Github contributions (5)
google-deepmind/acme
May 2020 - Jan 2023
A library of reinforcement learning components and agents
Role in this project:
ML Engineer
ML EngineerContributions:6 commits in 2 years 7 months
Contributions summary:George contributed to the `acme` reinforcement learning library by implementing new modules and improving existing ones. Their work includes adding `StochasticModeHead` and `ApproximateMode` modules within the `acme.networks` package. They also fixed a logic bug in `ExpQWeightedPolicy`, updating the code to use logits for more accurate probability calculations. Additionally, the user made improvements by incorporating configuration options such as a checkpoint TTL and making a hardcoded reward value configurable.
reinforcement-learningreinforcementagentsdeep-reinforcement-learning
google-research/google-research
Jul 2019 - Jan 2020
Google Research
Role in this project:ML Engineer & Data Scientist
ML Engineer & Data ScientistContributions:8 commits in 5 months
Contributions summary:George primarily contributes to the implementation and improvement of machine learning models within the Google Research repository. Their work involves implementing energy-inspired models, updating code to Python 3, and addressing various bug fixes and simplifications, specifically related to the LARS algorithm. The user also integrates and refines the use of TensorFlow datasets and implements convolutional neural networks and model variations. These contributions collectively demonstrate the user's proficiency in model development, data handling, and code maintenance within a research-oriented machine learning project.
googlemachine-learningai
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.