Neel Nanda
Senior Research Scientist at Google DeepMind
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolNeel Nanda is a Senior Research Scientist leading mechanistic interpretability work at Google DeepMind, focused on making advanced AI systems safe and understandable. With eight years of experience across DeepMind, Anthropic, and independent research, he builds and evaluates model architectures such as sparse autoencoders and probes transformer internals. He created TransformerLens, an influential open-source library for mechanistic interpretability, and has published interpretable analyses on phenomena like grokking and neuron discovery. A Cambridge mathematician grounded in Effective Altruism, Neel channels his research toward high-impact outcomes and donates 10% of his income to effective charities. He combines deep theoretical insight with practical engineering—mentoring junior researchers and releasing hundreds of open-weight models to broaden access to safety research. An unusual strength is translating complex mechanistic findings into reproducible tools and tutorials that let others explore model internals.
8 years of coding experience
3 years of employment as a software developer
BA, Mathematics, BA, Mathematics at University of CambridgeThe Latymer School
Github Skills (6)
transformers10
transformer-models10
pytorch10
machine-learning10
python10
natural-language-processing10
Programming languages (6)
Github contributions (5)
TransformerLensOrg/TransformerLens
Aug 2022 - Jan 2023
A library for mechanistic interpretability of GPT-style language models
Role in this project:
ML Engineer & Data Scientist
ML Engineer & Data ScientistContributions:5 releases, 89 reviews, 285 commits in 5 months
Contributions summary:Neel appears to be adding to and modifying a library designed for mechanistic interpretability of GPT-style language models. Their contributions include implementing and refining the model architecture (Embed, Unembed, PosEmbed, LayerNorm, Attention, MLP, TransformerBlock) and adding support for different models. A basic demo code for the model is added.
neelnanda-io/neelutils
Oct 2022 - Mar 2024
Contributions:13 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.