Reflection does not only emerge in SFT or RL stages but arises earlier during pre-training.

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

• Reflection-inducing directions emerge more clearly in higher layers (ℓ>5) for both models and datasetsfinding0.800
  Empirical observation about which network layers encode reflection-relevant information.
• Reflection is not merely a behavioral artifact of prompting but a phenomenon encoded in the model's activation space.claim0.783
  Central interpretive claim of the paper, supported by steering vector experiments.
• Reflective reasoning requires late-stage integration of semantic and reasoning signals, hence reflection-related directions emerge more clearly in higher network layers.claim0.779
  Interpretive claim about the locus of reflection in transformer architecture.
• When does the model initiate reflection during its reasoning process?question0.771
  First central research question motivating ReflCtrl investigation
• Introspective signals appear in middle layers but are suppressed by later post-training-shaped layers.finding0.766
  Mechanistic finding by Lindsey (2026) explaining how contemplative prompt may work: enables mid-layer introspection to reach output.
• Does ESR emerge from RLHF or does it exist in pretrained representations?question0.764
  Open question about developmental origin of ESR mechanisms
• A linear reflection direction exists in reasoning LLMs' latent representation space that governs self-reflection behaviorclaim0.764
  Core claim of ReflCtrl that a single direction captures and controls reflection
• We hypothesize ESR may emerge from RLHF training rather than existing in pretrained representationshypothesis0.760
  Open question about the developmental origin of ESR mechanisms