Tag
Anthropic's new J-lens reveals a hidden workspace inside Claude aligning with consciousness theory; other news include Tencent's Hy3 model, NVIDIA's compute-for-equity program, Microsoft layoffs, and Claude Fable 5 extension.
Anthropic's new research introduces 'J-lens,' a method to read Claude's internal activations before output, revealing a quiet workspace functionally similar to human global workspace theory. This allows detection of hidden reasoning, goals, and potential safety issues like prompt injections.
This paper introduces the Jacobian Lens (J-lens) and J-space to show that LLMs like Claude Sonnet 4.5 maintain verbalizable internal representations that function like a global workspace, causally used for flexible reasoning—demonstrated through intervention experiments.