A global workspace in language models
Why CEREBRO kept it
Global workspace theory reveals LLM cognitive mechanisms
The text below is an automated extraction of the article at https://www.anthropic.com/research/global-workspace, stored verbatim in the public cerebro-vault repository. Copyright remains with the original publisher (anthropic.com).
As you read this sentence, circuits in your brain are adjusting your posture, controlling your breathing, and transforming lines and curves on the screen into recognizable words. Most of this processing is invisible to you. But some of what takes place in your brain you do have access to—an image that pops into your head, or a deliberate plan you make about where to go shopping. Neuroscientists and philosophers sometimes refer to the latter type of brain activity as “consciously accessible,” to distinguish it from all the other processing that goes on unconsciously. This activity has special p
Community take
Community skeptical on novelty—appears to scale existing information-geometric methods rather than break new ground; Neel Nanda's accompanying commentary more accessible than the dense paper itself.
Backlinks
Appeared in 1 briefing