Introspective capacity may follow a simple monotonic scaling law across all concepts and architectures

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

• We hypothesize that introspective capabilities may scale with model size and architecture, including recurrence/looping that extends the integration windowhypothesis0.840
  Forward-looking prediction about whether early-layer introspection generalizes to larger models or recurrent architectures
• Introspective capacity scales with model size for some concepts, approaching near-perfect coupling in LLaMA-3.1-8Bclaim0.833
  Validated for wellbeing and interest; focus and impulsivity do not show consistent scaling
• This introspective capacity is highly unreliable and context-dependent in today's modelsclaim0.821
  A caveat qualifying the main claim.
• Introspective capacity is present from the first conversation turn, not requiring multi-turn context to emergeclaim0.813
  Three of four concepts show significant introspection at turn 1; rules out joint temporal drift as sole explanation
• Introspective capabilities have threshold effects requiring very large models; 70B models are barely on the threshold, and independent researchers lack access to larger models.claim0.807
  Practical bottleneck explaining why these phenomena are not widely studied.
• Either introspection is an emergent capability requiring larger scale, or more stringent controls are needed to test introspection in smaller modelsclaim0.807
  Alternative interpretations offered for why binary detection fails in Llama 3.1 8B but frontier models claim success
• Why does introspective capacity vary concept-by-concept and what mechanisms could stabilize it over time?question0.804
  Open question identified by the paper as direction for future work
• Is introspection an emergent property of scale, or do smaller open-weight models exhibit similar capabilities?question0.798
  Motivates comparison of Llama 3.1 8B results against Lindsey's frontier model findings