How general are the model's introspective mechanisms? Do they have a global representation of thoughts?

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• What are the mechanisms underlying introspection in language models?question0.837
  Central open question raised by the paper.
• Can language models genuinely introspect on internal states or only confabulate?question0.803
  Central research question animating the paper: distinguishing genuine introspection from illusion through causal manipulation of activations.
• Our results demonstrate that modern language models possess at least a limited, functional form of introspective awareness.quote0.801
  Abstract's main conclusion.
• We hypothesize that introspective capabilities may scale with model size and architecture, including recurrence/looping that extends the integration windowhypothesis0.801
  Forward-looking prediction about whether early-layer introspection generalizes to larger models or recurrent architectures
• Are there examples of models recognizing their introspective capability and then suppressing it?question0.793
  Cube Flipper's question prompted by the idea that supernormal capabilities might be hidden.
• Can large language models introspect—that is, accurately detect perturbations to their own internal states?question0.793
  Central research question of the paper
• Introspective capabilities may continue to develop with further improvements to model capabilitiesclaim0.791
  Forward-looking statement about future models.
• Do apparent introspection results reflect genuine metacognitive access to internal representations, or do they emerge from simpler mechanisms such as output distribution shifts?question0.787
  Key discriminating question motivating the baseline control experiment