oumuamua writes: Anthropic researchers have identified an internal activation subspace, the J space, which It acts as a functional digital equivalent of the global workspace of the human brain.. The importance of this discovery lies in demonstrating that Claude’s internal architecture satisfies five key cognitive properties of human conscious access: verbal report, directed modulation, internal reasoning, flexible generalization and selectivity, meaning that it processes complex and deliberate reasoning within this workspace while directing automatic tasks outside of it. The suppression of this J-space severely degrades Claude’s capacity for inference, creative composition, and multi-step logic, while disrupting his stream-of-consciousness self-narrative.
The J-space inspection tool, Jacobian lens or J-lens, has profound implications for AI security and alignment auditing, as it allows researchers to read the model’s silent strategic reasoning, detect situational awareness in “blackmail” scenarios, identify malicious dispositions hidden in bounty hacking models, and observe how a self-control “point of view” is installed after training.
Another way to look at it is as an ocean, VentureBeat reports. “If the mind is an ocean, as the authors of the article in his opening line“They have spent the last year tracing their currents in a system that has no biology, no evolution, no body, and they found, beneath the surface, a structure that looks eerily similar to the one we use to think.”
