framework
active
framework:jumprelu-saeJumpReLU SAE
Evaluated architecture.
Neighborhood — ranked by edge-count
Papers (5)
paper
- Do Sparse Autoencoders Capture Concept Manifolds?citesmentions
Datasets (1)
dataset
- Training data for L(n) curves.
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.
- Activation function not used; ReLU used instead.
- Baseline SAE architecture.
- Fig 6.
- Design justification.
- The individual, supposedly monosemantic directions learned by SAEs; argued here to fragment manifolds into disconnected pieces.
- SAE variant trained on ResNet18.
- Evaluated architecture.
- Evaluated architecture (Gao et al.).