Anthropic Discovers J-Space Internal 'Global Workspace' in Claude Model on July 7

According to Anthropic research announced on July 7, the Claude model has spontaneously developed an internal mechanism called J-space (Jacobian space), described as a 'global workspace' for silent reasoning. This neural activation pattern allows the model to think about concepts without articulating them, similar to human conscious thought. Using Jacobian techniques, researchers can identify and modify these activation patterns. Experiments show that disabling J-space weakens multi-step reasoning abilities but does not affect basic tasks or factual recall.
Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments