Researchers from Meta, MIT, and the University of Washington introduce Context Language Models (CLMs), a new approach that enables language models to manage and edit their own context rather than rely