Eliciting Latent Knowledge (ELK)

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

• latent reasoningconcept0.742
  Reasoning approach using learnable hidden embeddings.
• Post-Hoc Rationalization Elicitationmethod0.725
  Asking model to explain its own behavior after the fact when no chain-of-thought was available
• Latent Introspectionconcept0.725
  Pearson-Vogel et al.'s finding that models can detect prior concept injections; introspective signals exist in middle layers suppressed by post-training
• Connecting the Dots: LLMs Can Infer and Verbalize Latent Structure from Disparate Training Data (Treutlein et al. 2024)concept0.719
  Out-of-context reasoning work directly related to synthetic document fine-tuning experiments
• Friston, FitzGerald et al. (2016) — Active inference and learningconcept0.718
  Prior active inference paper providing detailed neurophysiological implementation of belief updates
• Stimulus-Elicited Intentionconcept0.716
  Zaadnoordijk and Bayne's category of intentional action; sticker-removal behavior induced by the self-prior corresponds to this
• That animal sentience is inferred on the basis of some likeness to humans is ultimately an intuition.claim0.713
  Claim that the basis for inferring animal sentience is intuitive, not empirical.
• Latent entitiesconcept0.712
  Entities that become visible as centers in a configuration (e.g., rectangles of white space around a dot) that were not present before.