Jul 6, 2026InterpretabilityA global workspace in language models

Anthropic researchers found that their language model Claude has a internal workspace, called the J-space, that allows it to think about concepts without outputting them. This workspace has properties similar to those of human conscious processing.

1sec.ai