We push the frontier of technical AI safety by fostering a global community of researchers.
We host the Sprints and the Lab. The Apart Sprints are regular 3-day events for engaging in object-level research on impactful topics within AI safety, such as Multi-Agent Risk, Model Evaluations and Interpretability. The Apart Lab invites top research teams from the Sprints to partake in a 3-6 month mentorship program towards publication.
Read more about the Apart Sprints' impact and the structure of the Apart Lab program.
We combine expertise from diverse fields including data science, physics, game development, neurotechnology, and machine learning, along with experience from academia, non-profits, startups, and industry, to tackle AI safety challenges. Feel free to reach out using the contact form at the bottom of the page.
Our advisors keep our core team on track to make valuable impact on the most important questions in AI safety.
Our research advisors provide the expertise and focus that ensure our research teams stay on course towards impactful and useful AI security and safety research.
