In-Context Neurofeedback: Can LLMs Control Their Internal Representations through Privileged Access?

A new study on arXiv (2609.00904v1) finds that large language models (LLMs) do not demonstrate reliable control over privileged internal representations when the control target is not inferable from the prompt, challenging a previous claim that LLMs can control their internal representations via neurofeedback. The authors argue that earlier results may rely on superficial mechanisms and that rigorous assessments of machine metacognition require privileged access.

arXiv:2609.00904v1 Announce Type: new Abstract: Whether large language models LLMs can control their own internal representations matters for both machine metacognition and AI safety. A recent study applied neurofeedback to LLMs and claimed that they can control their internal representations. However, the reported control may rely on superficial mechanisms rather than genuine internal access because the control targets in that study are not privileged, meaning that a third party can infer them from the prompt. We redesign the neurofeedback paradigm for LLMs so that the control target satisfies the privileged access requirement, which is closer to neurofeedback experiments in human cognitive neuroscience. Under this stricter setting, the models do not demonstrate reliable control over privileged internal representations, suggesting that previously reported control cannot exclude the possibility that it relies on superficial mechanisms. Our results indicate that rigorous assessments of metacognition in LLMs require evaluation methods that demand privileged access.