concept
active
concept:residual-persistenceresidual persistence
Emotion feature persistence above and beyond the persistence expected from high variance explained alone, computed by subtracting median variance-matched probe persistence
Neighborhood — ranked by edge-count
Papers (1)
paper
Methods (1)
method
- Variance-matched random probe controlimplementsControl method sampling random directions from top-k PC spaces matched to emotion probe variance, to isolate emotion-specific persistence
Concepts (1)
concept
- autoregressive persistenceextendsBaseline persistence of any probe direction arising from the autoregressive nature of LLMs, not specific to emotion content
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Proposed pathway flowing through layers at each position; calculates K/V values that feed horizontal information flow.
- The phenomenon that emotion feature activations remain elevated above baseline beyond local token bursts, measurable as long-range correlation
- Core logical paradox: if a species fails to change it dies; if it changes, it ceases to exist. Same applies to individuals.
- Core logical puzzle: if an agent does not change, it dies; if it changes, the self ceases to exist. Applies to all scales from organelles to evolutionary lineages.
- The property of emotion features maintaining elevated activation well beyond the local token context that triggered them
- The phenomenon where activating an emotion feature leads to subsequent below-baseline activation of that feature