By watching the J-space, we can see Claude silently perform reasoning steps in its head—noticing bugs in code, identifying images, and more.

This post discusses how observing Claude's internal activation patterns (J-space) reveals the model's hidden reasoning processes, such as detecting code bugs and analyzing images, even when these steps aren't explicitly shown in the model's output. The observation suggests Claude performs silent intermediate reasoning that can be made visible through internal state analysis.

Source: Yohei Nakajima (X) | 2026-07-06