The J-Space: How I Learned To Read An LLM's Mind

Anthropic researchers have identified a 'J-space' within Claude, a set of internal reasoning steps that precede the model's final word choice. Although this space holds only a small fraction of a concept's representation, disabling it causes Claude's multi-step reasoning to collapse, while fluent speech and simple recall remain intact. The discovery offers a rare glimpse into the internal workings of large language models.

Last week, Anthropic published a paper on a discovery they made regarding Claude's internal reasoning they call the J-space . It seems to be the steps that Claude works through before it commits to a final word: the closest thing we've come to seeing the inside of an LLM's "brain." Although it holds a median of only 6–7% of a concept's representation inside the model and never more than about a tenth of the model's activity at any layer, if you switch it off, Claude's multi-step reasoning collapses to almost nothing. Fluent speech and simple recall remain intact. I quite liked the video that Anthropic released https://www.youtube.com/watch?v=rKV5JcALQoQ with the research post https://www.anthropic.com/research/global-workspace . It's a long research post though, so I thought I'd write a summary of the parts I considered the coolest and how I interpreted the mathematics of the J-space. This is a preview of a post from my blog. Read the full post with the interactive charts →