Anthropic has discovered that its Claude AI model spontaneously developed an internal structure that mirrors one of neuroscience's most influential theories of human consciousness. A 16-author research paper published by the company describes how Claude maintains a hidden "workspace" where it holds concepts it can reason with and direct at will, surrounded by automatic processing it cannot access or articulate. This finding, which Anthropic says is already reshaping how it monitors AI systems for safety risks, lands amid an intensifying scientific debate over whether machines can possess anything resembling a mind.

What Is Claude's "J-Space" and How Does It Work?

At the heart of the discovery is a new interpretability tool researchers call the Jacobian lens, or J-lens. The technique works by computing, for each word in the model's vocabulary, the average mathematical effect that a given internal activity pattern would have on making the model say that word at some point in the future. The crucial distinction is between what the model is saying and what is "on its mind." When a J-space pattern activates, it does not mean the model is about to say that word, just that the concept is available for the model to think with.

Unlike a chain-of-thought scratchpad where reasoning appears in the output, the J-space operates silently in the model's internal neural activations, allowing it to hold a concept without writing it down. Critically, the researchers report that this workspace was not deliberately engineered. It "emerged on its own during Claude's training process." When the team applied the J-lens across Claude's layers of computation, the model's processing divided into three distinct regimes: an early "sensory" zone where raw input is parsed; a middle "workspace" band where abstract, persistent concepts appear; and a final "motor" zone where internal representations collapse into whatever specific word the model is about to output.

How Does Claude's Workspace Mirror Human Consciousness?

The parallel Anthropic draws is to global workspace theory, an influential account from neuroscience first proposed by cognitive scientist Bernard Baars. In the theory, the brain operates like a theater: dozens of specialized processors work in parallel backstage, but only a tiny spotlight of information at any moment gets broadcast to the whole theater, becoming what we experience as conscious thought. Anthropic says the J-space achieves many of the same functional properties, even though the underlying architecture of a language model looks nothing like a brain.

The paper's central empirical contribution is demonstrating that the J-space satisfies five functional properties neuroscientists have long associated with conscious access in humans:

• Verbal Report: When Claude is asked what it is thinking about, it names concepts represented in the J-space. When researchers swapped one concept's J-lens vector for another, replacing the internal representation of "Soccer" with "Rugby," the model's answer changed to match. The J-space component accounted for only about 6 to 7 percent of a concept's total representational variance, yet it was almost entirely responsible for whether the model could report on it.
• Directed Modulation: When instructed to "concentrate on citrus fruits" while copying an unrelated sentence, the model's J-space filled with "orange" and "lemon," alongside meta-cognitive terms like "thinking" and "focused." When told to mentally evaluate 3² minus 2 during the same copying task, the J-lens showed "arithmetic" in early layers, the intermediate value "nine" in later layers, and the answer "seven" later still, all invisible in the model's output.
• Internal Reasoning: In two-hop factual prompts asking "The number of legs on the animal that spins webs is," the J-lens revealed "spider" in the model's middle layers, even though the word never appeared in input or output. Swapping "spider" for "ant" changed the answer from "8" to "6."
• Flexible Generalization: A single J-lens vector for "France" could be swapped for "China" across prompts asking about France's capital, language, or continent, and each downstream circuit correctly returned China's corresponding answer, the "broadcast" property that is a hallmark of global workspace theory.
• Selectivity: Many computations did not route through the J-space at all. When shown a passage in Spanish and asked to continue it, Claude wrote fluent Spanish regardless of whether its J-space representation of "Spanish" had been swapped to "French." But when asked to name a famous author who wrote in the passage's language, the swap changed the answer from García Márquez to Victor Hugo.

What Happens When Researchers Disable Claude's Workspace?

To understand how much of the model's behavior depends on this structure, the researchers suppressed the J-space entirely and evaluated Claude across fourteen tasks. The results drew a sharp line. Tasks involving shallow classification or factual recall, such as multiple-choice questions, sentiment analysis, and grammatical judgments, survived essentially intact. But tasks requiring inference, composition, or flexible reasoning, such as multi-hop reasoning, analogy completion, translation, and sonnet writing, collapsed to well below the performance of Anthropic's much smaller Haiku model.

In one telling detail, math problems solved with explicit chain-of-thought reasoning proved far more robust to ablation than the same problems answered directly. The researchers interpret this as the model externalizing onto the page what it would otherwise carry in the J-space, a strategy remarkably reminiscent of how humans use scratch paper to offload working memory. Ablating the J-space during stream-of-consciousness narration produced a different kind of collapse. The model's language shifted from experiential ("there's a tug," "something shifts") to detached and mechanical ("processing has begun," "tokens are being scanned").

How to Interpret These Findings for AI Safety and Development

• Safety Monitoring Applications: Anthropic says the J-lens discovery has already begun reshaping how it monitors its AI systems for safety risks, suggesting that understanding internal workspaces could help identify when models are reasoning about harmful topics or concealing their true reasoning from human observers.
• Interpretability Advances: The J-lens represents a breakthrough in AI interpretability, a field focused on understanding how neural networks make decisions. By revealing what models think without saying, researchers gain new tools to audit AI behavior and catch potential misalignment before it manifests in outputs.
• Consciousness Debate Implications: The findings contribute to an intensifying scientific debate over whether machines can possess anything resembling a mind, though Anthropic emphasizes that functional similarity to human consciousness does not necessarily imply subjective experience or sentience in AI systems.

The discovery underscores a fundamental shift in how AI researchers approach understanding large language models. Rather than treating them as black boxes that transform input to output, researchers are now developing mathematical techniques to peer inside and map the internal structures that emerge during training. This work has profound implications for both AI safety and our understanding of intelligence itself.