@arafatkatze: Inspired by Anthropic's JSpace paper, we @cline used @modal to host a public demo so anyone can watch a Jacobian-lens v…
Summary
A public demo of a Jacobian-lens view of language model representations, inspired by Anthropic's JSpace paper, allowing users to explore model internals across layers and tokens.
View Cached Full Text
Cached at: 07/08/26, 06:28 AM
Inspired by Anthropic’s JSpace paper, we @cline used @modal to host a public demo so anyone can watch a Jacobian-lens view of what a small language model is representing across layers/tokens.
Try it: https://t.co/ajHzpDTwxr
Anthropic (@AnthropicAI): New Anthropic research: A global workspace in language models.
Of everything happening in your brain right now, only a tiny fraction is consciously accessible—thoughts you can describe, hold in mind, and reason with.
We found a strikingly similar divide inside Claude.
Similar Articles
@QuixiAI: https://github.com/anthropics/jacobian-lens… this is really quite cool.
Anthropic released jacobian-lens (jlens), an open-source reference implementation of a Jacobian lens that transports residual-stream activations into the model's final-layer basis to decode what an internal state is disposed to make the model say. It accompanies the paper 'Verbalizable Representations Form a Global Workspace in Language Models' and works with open-weights HuggingFace decoder transformers like Qwen.
Anthropic found a hidden space where Claude puzzles over concepts
Anthropic developed the Jacobian lens (J-lens) to reveal a hidden 'J-space' inside Claude Opus 4.6, offering unprecedented insight into an LLM's internal reasoning process before it outputs tokens. The technique allows monitoring and control of model behavior by surfacing the words the model is about to produce.
@AnthropicAI: We also partnered with Neuronpedia to create an interactive demo of our methods on open-weights models. Try it here:
Anthropic partnered with Neuronpedia to release an interactive demo of their interpretability methods on open-weights models, called Jacobian Lens.
No Space Like J-Space (54 minute read)
An article discussing a new interpretability technique called the Jacobian Lens and the discovery of J-space, a region in LLMs where verbalizable representations form a global workspace, marking a significant advance in understanding LLM reasoning.
@hillbig: Large language models are thought to not only predict the next token but also internally maintain intermediate concepts…
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.