Goodfire

Our mission is to advance humanity's understanding of AI by examining the inner workings of advanced AI models (or "AI Interpretability"). As an applied research lab, we bridge the gap between theoretical science and practical applications of interpretability to build safer and more reliable AI models.

In-Context Learning &

In-Context Learning & "Model Systems" Interpretability (Stanford lecture 3) - Ekdeep Singh Lubana
Computational Motifs (Stanford lecture 2) - Jack Merullo

Computational Motifs (Stanford lecture 2) - Jack Merullo
Causal Mechanistic Interpretability (Stanford lecture 1) - Atticus Geiger

Causal Mechanistic Interpretability (Stanford lecture 1) - Atticus Geiger
Paint with Ember

Paint with Ember
Goodfire Ember Quickstart

Goodfire Ember Quickstart
Goodfire Ember Auto Steer Demo

Goodfire Ember Auto Steer Demo
Editing Llama to be conscious - Goodfire's research preview sneak peek

Editing Llama to be conscious - Goodfire's research preview sneak peek
The power of feature steering - Goodfire's research preview sneak peek

The power of feature steering - Goodfire's research preview sneak peek
Introducing Goodfire's research preview

Introducing Goodfire's research preview