Qwen's J-Space - Anthropic's discovery of an internal model Global Workspace

Why it matters

Claude model activity is surfacing — worth tracking for capability changes, ecosystem impact, and availability.

AI summary

Anthropic's new research explores the internal thought processes of AI models before generating responses. They've released

Anthropic published research today into what a model is thinking behind the scenes while it is deciding what to actually write.

More importantly, they released the J-Space lens code and their partner put together a demonstration of Qwen 3.6 27B J-Space .

TopicsClaudeModel release

Keywords#anthropic#discovery#workspace#internal#global#model

Single source, no cross-check yet