claim
active
claim:current-sae-architectures-directly-motivated-by-manifold-learning-algorithms-face-an-impossibility-resultCurrent SAE architectures directly motivated by manifold learning algorithms face an impossibility result
Related to point-based simplicial limit.
Source paper
extracted_from(2026) · Usha Bhalla · Thomas Fel · Can Rager · Sheridan Feucht +8
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Justifies focusing on direction-based SAEs.
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.
- Theoretical framework answer.
- Features act as local patch detectors.
- Feature cluster examples from the UMAP.
- Core concern.
- Manifold-level descriptions recover overarching semantic structure that SAE features miss.claim0.783Positive claim that geometric descriptions retain the conceptual coherence lost in atomized feature decompositions.
- Restricted to specific manifolds, different SAEs show strong alignment of induced coordinate systemsfinding0.778Geometry is shared even if features are not.
- A promising property for interpretability analysis off-distribution.