emotion feature persistence

The phenomenon that emotion feature activations remain elevated above baseline beyond local token bursts, measurable as long-range correlation

Neighborhood — ranked by edge-count

Papers (1)

paper

Persistence and Introspection of Emotion Features
introduces

Concepts (4)

concept

Anti-Persistence of Emotion Features
related_to
The phenomenon where activating an emotion feature leads to subsequent below-baseline activation of that feature
bursty emotional dynamics with long tail
extends
The observed pattern of fast initial spike in emotion activation followed by slow decay to a persistent elevated baseline, characterizing emotion feature dynamics
internal emotional state
associated_with
The possibility of a stably encoded, causally active emotional state within LLMs, as distinct from token-by-token semantic content
PCs of the emotion space and persistence
extends
Analysis showing that lower-rank (more central) PCs of emotion feature activations are more persistent than higher-rank (noisier) PCs

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

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

Emotional State Persistenceconcept0.849
The property of emotion features maintaining elevated activation well beyond the local token context that triggered them
To what extent is emotion feature persistence driven by genuine internal emotional state versus autoregressive conversational context dynamics?question0.829
Core open question the paper raises but does not fully resolve
If persistence is genuinely related to emotion features, lower PCs of the emotion space (more central, less noisy) should be more persistent; if it is an artifact, noisier PCs should have similar persistence.hypothesis0.817
Falsifiability test built into the PC analysis design
Whether observed persistence reflects a genuine lingering emotion-like state or merely persistent conversational context that produced the emotion-relevant activationquestion0.796
Core unresolved confound the paper acknowledges but cannot rule out
residual persistenceconcept0.790
Emotion feature persistence above and beyond the persistence expected from high variance explained alone, computed by subtracting median variance-matched probe persistence
Emotion features are not strictly locally scoped; they are bursty with a long tail of slow change persisting over 100 tokens.claim0.786
Main conclusion about the temporal dynamics of emotion features
Persistent conversational context that produced emotion-relevant activation is a plausible driver for the observed persistence results.claim0.785
Acknowledged alternative explanation that the paper does not rule out
Emotion refers to a state concept, so stateful representations in general may be more persistent across tokens.claim0.783
Interpretive hypothesis offered to explain why emotion features are more persistent