Teams at Goodfire

Conduct original research on interpretability of large AI models, prototype visualization and manipulation techniques, collaborate with engineers to productionize tools, publish and open-source results, and help define research direction focused on mechanistic understanding and model steering.