BREAKING
Anthropic finds 'J-space' in Claude
Reading Claude's silent thoughts
Five functional properties
1Reportability
2Deliberate control
3Causal in reasoning
4Flexible use
5Limited scope
J-space vs chain-of-thought
J-space
Internal activation patterns
Editable to redirect reasoning
Emerged during training
Chain-of-thought
Expressed as visible text
Read as output
Explicit reasoning steps
A tool, not proof of consciousness
Rewriting AI's inner workings
AI NEWS BLITZ
Anthropic says Claude spontaneously grew an internal workspace called J-space.