Distribution-Level Prompt

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

• The mode of a distribution-level prompt approximates the diverse distribution learned by the base model during pretrainingclaim0.804
  Supported by empirical comparison showing VS achieves KL divergence of 0.12 from pretraining distribution vs. 14.89 for direct prompting
• Distributed representationconcept0.775
  Idea that information is spread across many neurons; superposition is a subtype.
• Prompt-Label Baselinemethod0.769
  Conditional generation on explicit Big Five labels using per-dimension descriptors; used as inference-time baseline
• Preferred Distributionconcept0.754
  In active inference, the distribution over goal states; here replaced by the learned self-prior rather than a hand-specified prior
• Natural Distribution of Representationsconcept0.750
  The distribution of latent representations produced by the model under unperturbed inputs
• Declarative-style deployment prompts (e.g., 'INFO: Not evaluation.') reduce type hint rate to 34-37%, approaching deployment levelsfinding0.741
  Weakens overall setup by showing some prompts can lower type hints, but does not invalidate core steering result.
• Different prompt types collapse to different modes: instance prompts collapse to a single prototypical response, while distribution prompts collapse to a representative high-entropy sampleclaim0.739
  The theoretical mechanism explaining why VS works despite mode collapse remaining operative
• Generate and Testframework0.737
  Abstract iterative scheme underlying Darwinian evolution, GP, and Dennett's tower of learning.