Prompt Brittleness and Long-Context Dilution

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

• Distribution-Level Promptconcept0.718
  A prompt framing requesting a representative sample from a distribution rather than a single instance, which is the key insight behind VS
• Contrastive system prompt completionsconcept0.714
  Training method for probes: generate completions under opposing system prompts to induce positive and negative poles of a concept
• Task difficulty moderates whether CoT is performative or genuine: easy recall questions show performative CoT, difficult multihop questions show genuine reasoningclaim0.712
  Task difficulty as the key variable distinguishing the two modes of CoT identified in the paper
• Persistent conversational context that produced emotion-relevant activation is a plausible driver for the observed persistence results.claim0.708
  Acknowledged alternative explanation that the paper does not rule out
• Practical context length limitations in language models lead to forgetting outside the window, constraining coherence over time.claim0.708
  Claim about engineering constraint reinforcing the theoretical no-order result
• Random word prefix prompts show emergence patterns similar to no-prompt, suggesting prompt length alone does not shift truth geometry.claim0.706
  Control experiment ruling out token-count as the cause of truth geometry shifts.
• Persistent conversational context that produced emotion-relevant activations is a plausible driver of observed persistence resultsclaim0.706
  Authors' caveat that conversational context persistence rather than internal emotion state persistence could explain findings
• Prompt and context design are cognitive-control operations: they toggle latent competencies rather than teaching the model from scratch.claim0.705
  Assertion about the nature of prompt engineering.