hypothesis
active
hypothesis:using-p-norms-with-p-1-would-alleviate-echo-featuresUsing p-norms with p<1 would alleviate echo features
Mitigation proposal.
Source paper
extracted_fromNeighborhood — ranked by edge-count
Papers (1)
paper
Findings (1)
finding
- Seen as in other 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.
- L1-of-norms is preferable for loss comparability and exposing layer- or model-specific featuresclaim0.733All experiments used L1-of-norms.
- Task-dependent optimal sparsity.
- First of three operational criteria for identifying consciousness phenomena in LLM representations.
- Sparsity penalty weighting by sum of per-layer decoder norms, enabling comparability with per-layer SAEs.
- Table 3.
- Concurrent work result showing emergent misalignment occurs in small models
- Footnote 2.