The Inversion Lab advances AI Safety through interpretability research. We build methods to understand the behavior and inner workings of AI systems, and use these insights to mitigate risks when AI agents interact with the world and each other.
Learn more
Contact us via galke@imada.sdu.dk