Contrasting No Reflection with Triggered Reflection (µ(0→2)) provides a stronger reflection signal than contrasting Intrinsic with Triggered Reflection (µ(1→2)).

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

• Triggered Reflection with 'Alternatively' achieves accuracy .684 on gsm8k_adv for Gemma3-4B-ITfinding0.805
  Highest single-instruction accuracy result in the paper.
• Reflection is not merely a behavioral artifact of prompting but a phenomenon encoded in the model's activation space.claim0.784
  Central interpretive claim of the paper, supported by steering vector experiments.
• Enacted reflection may correspond to silent mid-layer processing; described reflection to the motor impulse of concepts leaking through to output.claim0.778
  Mechanistic analog connecting Lindsey's layer-localized findings to the scorer's enacted/described distinction
• Clear accuracy stratification across three reflection levels on cruxeval_o_adv: Triggered (.065/.247) > Intrinsic (.040/.133) > No Reflection (.017/.051) for Qwen2.5-3B/Gemma3-4B-ITfinding0.766
  Core empirical result validating the three-level reflection framework on code reasoning.
• Reflection-inducing directions emerge more clearly in higher layers (ℓ>5) for both models and datasetsfinding0.762
  Empirical observation about which network layers encode reflection-relevant information.
• Higher reflection frequency correlates with lower accuracy partly because more reflections are generated on difficult questionsclaim0.760
  Author's interpretation of the negative correlation between reflection rate and accuracy observed in Fig. 5
• Activation steering of reflection has dual-use implications: it can enhance reflection as a defense mechanism, but malicious actors may also use it to inhibit reflection to facilitate jailbreaks.claim0.756
  Applied dual-use conclusion drawn from the paper's findings.
• µ is an applicative homomorphism: µ(pure a) = pure a and µ(imf <*> imx) = µ imf <*> µ imx.claim0.755
  Result for Image applicative specification.