Lee Sharkey
Principal Investigator
London, England, United Kingdom
Summary
👤
Senior🎓
Top SchoolLee Sharkey is a principal investigator and interpretability researcher with nine years' experience unraveling what neural networks really learn to ensure safe deployment. Based in London and trained at ETH Zurich and Cambridge, he combines expertise in high-dimensional data analysis, visualisation, and mechanistic interpretability across PyTorch, TensorFlow, JAX and Keras. He has co-founded a research startup, published multiple influential papers on neural network geometry and RL agent analysis while at Conjecture, and continues to lead applied interpretability work at Goodfire. His background in public health and global organizations gives him an unusual systems-level perspective on risk and robustness in AI. Colleagues know him for pragmatic methods—like layer decomposition techniques—and for bridging rigorous research with deployable tools.
9 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD (discontinued), Human and Machine Cognition Lab & Neurocognitive modelling group, Doctor of Philosophy - PhD (discontinued), Human and Machine Cognition Lab & Neurocognitive modelling group at University of TübingenMaster of Science - MS, Neural Systems and Computation, Master of Science - MS, Neural Systems and Computation at ETH ZurichMethodist College BelfastBachelor of Arts (BA), Pre-Clinical Medical Sciences and Neuroscience, Bachelor of Arts (BA), Pre-Clinical Medical Sciences and Neuroscience at University of CambridgeMSc in Data Analytics, with Distinction, MSc in Data Analytics, with Distinction at University of Glasgow
Github Skills (2)
pytorch8
openai8
Programming languages (3)
Github contributions (5)
jbkjr/train-procgen-pytorch
Mar 2021 - Jun 2022
Pytorch implementation on OpenAI's Procgen ppo-baseline, built from scratch.
Contributions:447 commits, 226 pushes, 20 branches in 1 year 2 months
openaipytorch
leesharkey/deep_attractor_network
Nov 2020 - Apr 2022
Contributions:2 pushes in 1 year 5 months
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.