Liv GortonMember of Technical Staff · AnthropicResearcher↗
Researcher · Founder
Christopher Olah
AnthropicCo-Founder and Interpretability Research Lead
Olah pioneered modern neural-network interpretability, from feature visualization[1] and the Distill journal to mechanistic interpretability, which seeks to reverse-engineer a model's learned computations.[2] As an Anthropic co-founder he has built this into a major empirical program for understanding frontier models.
Career & education
Current roleAnthropic2021-presentCo-Founder and Interpretability Research Lead
Anthropic2021-presentPreviously
OpenAI2018-2020Member of Technical Staff
Google Brain2015-2018
OpenAI2018-2020Google Brain2015-2018Work and education history is primarily focused on AI-relevant roles and may not be comprehensive.
Citation Trend
Total citations to scholarly work, including research papers, technical reports, books and theses, as recorded by Google Scholar. Each point is a dated snapshot.
citations129,631As of Nov 1, '26
citations+3,498Since last snapshot, Oct 1, '26
Selected contributions
Concrete Problems in AI Safety
Outlined five practical research problems for preventing accidents in AI systems.
Constitutional AI: Harmlessness from AI Feedback
With Yuntao Bai, Saurav Kadavath, Sandipan Kundu, Amanda Askell, Jackson Kernion, Andy Jones, Anna Chen and others ·Trained AI assistants to critique and improve their answers using a written set of principles.
Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
With Yuntao Bai, Andy L. Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain and others ·Training AI assistants to be more helpful and less harmful using human feedback.
References
- Feature Visualizationdistill.pub
Original 2017 Distill article lists Chris Olah as first author and develops methods for visualizing what neural networks respond to.
- A Mathematical Framework for Transformer Circuitstransformer-circuits.pub
Original Anthropic research credits Chris Olah as author and corresponding author; defines mechanistic interpretability as reverse engineering model computations.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.
Reviewed August 10, 2026 · Checked for accuracy and new information.
