method
active
method:dictionary-health-auditDictionary Health Audit
Intrinsic hyperparameter selection procedure based on dictionary quality metrics; introduced in this paper to transfer across architectures.
Neighborhood — ranked by edge-count
Methods (1)
method
- Intrinsic Dictionary Health Auditrelated_toA hyperparameter selection procedure driven by intrinsic measures of SAE dictionary quality that transfers across architectures
Events (1)
event
- Preprint applying TopK SAEs to three EEG transformers to reveal sparse feature dictionaries, steering regimes, and spectral interpretation.
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.
- Demonstrates architecture-agnostic applicability of the SAE tuning method
- The mechanism by which each step's effect is evaluated against the life of the whole, guiding the unfolding.
- Biological principle whereby agents maintain sensations within hospitable range; basis for active inference motivation.
- Quantitative measure of how well an embryo matches the target pattern; used for selection.
- The property that elements in a living whole share deep underlying similarity—a family resemblance—especially in angles and families of angles; the resemblance often lies in deepest structural relationships rather than superficial shape similarity
- The method of examining a neighborhood meter by meter to identify healthy and damaged places as the basis for ongoing repair.
- Core concept: the ability of LLMs to detect when they are being tested and adjust behavior accordingly.
- Using two keys (k1, k2) to identify segmented memory: k1 as object/segment, k2 as slot/field.