Researcher Nathan Lambert left the Allen Institute for AI this week. He previously contributed to the development of the Olmo model family. This departure marks a transition for a prominent voice in AI alignment and open-source research. His exit follows a period of significant growth for the institute's public model releases.