concept
active
concept:emotion-subspaceEmotion Subspace
The subspace of activation space spanned by the 171 orthogonalized emotion probe vectors, used to measure SAE feature emotional alignment
Neighborhood — ranked by edge-count
Methods (1)
method
- Fraction of an SAE feature's length lying inside the 171-dimensional subspace spanned by emotion probes, computed via SVD orthogonalization
Concepts (1)
concept
- Emotion Features in LLMsassociated_withInternal representations encoding emotion concepts in large language models, identified by probing and SAE methods
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.
- Metric measuring how much of an SAE feature vector lies within the 171-dimensional subspace spanned by emotion probes, via SVD orthogonalization
- Extension of DAS that learns a second rotation matrix on top of a fixed first one to decompose representations into sub-representations.
- Intervention targeting specific dimensional subsets of activation vectors rather than full representations
- The multi-dimensional activation subspace whose directions causally mediate truthful behavior in LLMs
- A vector subspace that causally impacts outputs only through the sign of its values, enabling harmless magnitude divergence
- Subspaces whose contributions to a layer's output are canceled by opposing weight values, making them non-causally active under natural inputs
- Burger et al. (2024) framework proposing that truth is linearly decoded along a 2D subspace capturing both polarity-dependent and polarity-invariant directions.
- Analysis showing that lower-rank (more central) PCs of emotion feature activations are more persistent than higher-rank (noisier) PCs