concept
active
concept:linear-representation-of-featuresLinear Representation of Features
The central object of study — the idea that a concept like truth is encoded as a direction in the LLM's latent space
Neighborhood — ranked by edge-count
Papers (1)
paper
Concepts (3)
concept
- Linear representationrelated_toThe idea that features are encoded as directions in activation space.
- Linear Representation of Conceptsrelated_toThe established finding that transformer LLMs encode many interpretable concepts as linear directions in activation space
- Truth Direction in LLM Latent Spaceassociated_withA specific direction in an LLM's residual stream that encodes the truth or falsehood of factual statements
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.
- The hypothesis that models internalize concepts as approximately linear directions in representation space; used to interpret MDS injection behavior
- The finding that interpretable concepts including character traits are encoded as linear directions in transformer residual streams
- Premise that models represent concepts as features.
- Activations are sums of sparse features living in subspaces.
- Open question A.1.1a.2.
- The sequential, continuous order of text, often challenged by diagrammatic branching.
- Interpretable concept representations hypothesized as basic units of computation.
- Fraction of a dimension given to a feature.