concept
active
concept:sense-vectors

Sense Vectors

Vectors acquired during pretraining in Backpack LMs that have a multiplication effect on model generation

Neighborhood — ranked by edge-count

Frameworks (1)

framework
  • LM architecture with sense vectors showing multiplication effects, illustrating custom intervention in pyvene

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.

  • steering vectorsconcept0.813
    A method for modifying model behavior by adding perturbation vectors to activations, used here to try to reduce eval awareness.
  • Emotion vectorsconcept0.800
    Linear representations of emotion concepts extracted from activations.
  • concept vectorconcept0.795
    Computed directional vector in activation space representing a specific concept, used for injection experiments
  • Attribute-aligned shifts added to the residual stream to steer LLM generation toward desired personality traits
  • Persona Vectorsmethod0.780
    Directions in activation space encoding a model's dispositional self-presentation, used as evidence for self-modelling.
  • Function Vectorconcept0.779
    Type of steering vector enabling zero-shot task execution, cited from Todd et al. 2024
  • Used to extract trait and style directions.
  • Prior framework for monitoring and controlling character traits in LLMs via activation directions; this paper extends it to 275 roles