A global workspace in language models
ai
Anthropic researchers published a new paper yesterday titled "A Global Workspace in Language Models." The research investigates how workspace mechanisms might function within large language models. The concept of "global workspace" originates from cognitive science, where it describes how information becomes available across different processing systems. The full paper is available on Anthropic's website.
Source: https://www.anthropic.com/research/global-workspace
Listen to this story
Hear this and more stories in a personalized audio briefing.
Open The Chonkerton