Saying that LLMs cannot introspect or cannot introspect on what they were doing internally while generating or reading past tokens in principle is just dead wrong. The architecture permits it.

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

• Transformer architecture permits introspection; saying LLMs cannot introspect on past internal states is wrong.claim0.865
  Janus's central claim that the architecture enables introspection, though usage in practice is a separate question.
• So at any point in the network, the transformer not only receives information from its past... but also has causal influence over its future processing. So, saying that LLMs cannot introspect... is incorrect.quote0.858
  Core summary of Janus' position on autoregressive recurrence enabling introspection.
• Do LLMs leverage architectural capacity for introspection on internal computations and prior token generation?question0.854
  Central empirical question separating architectural possibility from actual model behavior; gates introspection research.
• LLM introspection on internal computations is architecturally permitted; whether models leverage this is an empirical question.claim0.839
  Core claim directly challenged by prior work denying introspection; forms foundation for Koan Battery introspection studies.
• LLMs can compute meaningful functions over perturbations to their internal states, establishing introspection as a real but layer-dependent phenomenonclaim0.827
  Primary positive claim of the paper, grounded in strength comparison and localization results
• How are LLMs actually leveraging the architectural degrees of freedom for introspection in practice?question0.816
  Janus notes that while architecture permits introspection, it is a separate question how models use it.
• "Our findings demonstrate that LLMs can compute meaningful functions over perturbations to their internal states, establishing introspection as a real but layer-dependent phenomenon that merits further investigation."quote0.812
  Central thesis statement of the paper
• The objection that feedforward networks cannot introspect is a cultural myth; autoregression provides recurrence across tokens.claim0.799
  Antra's rebuttal to a common criticism; backed by Janus' information flow diagram.