framework
active
framework:nested-saeNested SAE
Hybrid SAE jointly training turn-averaged inner and per-token outer partitions.
Neighborhood — ranked by edge-count
Papers (1)
paper
- Turn-Averaged SAEs for Feature Discovery and Long-Context Attributionintroducesmentions
Frameworks (2)
framework
- Paper's main proposal: SAE trained on the mean activation across a turn.
- Matryoshka SAEsextendsNested dictionary approach that inspires the nested loss.
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.
- Hybrid with separate turn-averaged and residual per-token SAEs.
- Evaluated architecture.
- Evaluated architecture.
- The individual, supposedly monosemantic directions learned by SAEs; argued here to fragment manifolds into disconnected pieces.
- Interpretable features extracted by sparse autoencoders used as steering targets in this study
- Evaluated architecture (Gao et al.).
- Interpretability method criticized in this paper for shattering manifolds into atomic pieces, obscuring overarching semantic structure.
- Internal concepts a model uses, extracted as human-interpretable components.