question
active
question:how-true-is-the-linear-representation-hypothesisHow true is the linear representation hypothesis?
Open question A.1.1a.2.
Source paper
extracted_from(2025) · Lee Sharkey · Bilal I. Chughtai · Joshua Batson · Jack W. Lindsey +25
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Position on LRH.
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
- Activations are sums of sparse features living in subspaces.
- Hypothesis that information may be encoded in arbitrary non-linear subspaces of a neural network
- Authors' overall conclusion from number of interpretable features, activation-level correspondence to intensity, sensible logit weights, and interference weights
- The idea that features are encoded as directions in activation space.
- Interpretive synthesis of DIM and cone intervention successes
- Speculative interpretation of breakdown.
- Machine-learning hypothesis (Huh et al. 2024) proposed as a candidate framework to unify with the Platonic space model.