concept
active
concept:concept-vectorconcept vector
Computed directional vector in activation space representing a specific concept, used for injection experiments
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.
- Procedure extracting concept vectors as difference of mean activations between concept-exemplifying and baseline/negative sentences
- The spatial/geometric organization of conceptual structure within neural network representations; central to the paper's thesis.
- Method for obtaining concept vectors by subtracting activations from two contrasting prompts.
- Central entity of Jackson's framework: a structure invented to give coherent account of immediate consequences of actions; the building block of software design
- Type of steering vector enabling zero-shot task execution, cited from Todd et al. 2024
- How a neural network encodes a semantic concept internally, argued to be better captured by manifolds than by atomic features.
- Probabilistic framework formalizing concept-specific subspaces for targeted steering in generative models.
- Vectors acquired during pretraining in Backpack LMs that have a multiplication effect on model generation