Features may not be strictly one-dimensional objects; higher-dimensional feature manifolds may exist in model representations

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Geometry of features matters for representation quality.claim0.803
  General principle supported tangentially by covariance pooling work; relates to feature co-occurrence structure.
• Feature Manifoldsframework0.801
  Hypothesized extension of superposition where features may be higher-dimensional manifolds rather than 1D directions
• Feature universality across independently trained models suggests features have some existence beyond individual modelsclaim0.780
  Authors take agnostic position on ontological status but universality evidence pushes toward features being real
• Lack of rigorous cross-model comparison demonstrating that specific named features (not just correlated ones) form across architecturesquestion0.777
  Explicitly identified research gap: anecdotal evidence exists but rigorous characterization is absent
• Representational abstraction of truth may emerge more clearly with model scaleclaim0.774
  Interpretation of weaker PCA separation and lower ASR in smaller models
• The two-dimensional subspace reported by Burger et al. (2024) seems to reflect a stage of transition in the model's processing, rather than a universal property of truth directions.quote0.773
  Load-bearing interpretive claim about the layer-specificity of Burger et al.'s finding.
• We hypothesize that representation geometry drives model behavior — the geometric structure of internal representations causally shapes what models do externally.hypothesis0.769
  The causal hypothesis motivating the use of causality (intervention) as the lens connecting representation and behavior geometry.
• Features can be used to steer large models.claim0.768
  Clamping feature activations causally alters model behavior in interpretable ways.