concept
active
concept:ultralow-density-clusterUltralow Density Cluster
Cluster of autoencoder features with extremely low activation density (~1e-7) that are generally not interpretable and appear to be training artifacts
Neighborhood — ranked by edge-count
Methods (1)
method
- Feature Density Histogramassociated_withLog-scale histogram of feature firing rates used as proxy for autoencoder quality during hyperparameter tuning
Concepts (1)
concept
- Feature Densityassociated_withFraction of training tokens on which a given feature has nonzero activation; used as proxy metric for autoencoder quality
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.
- Mode in feature density histogram around 1e-5 corresponding to interpretable features, contrasted with ultralow density cluster
- Third cluster of the four-cluster grouping, containing properties 3, 2, 8.
- Cluster 2 (4-cluster): LOCAL SYMMETRIES, THE VOID, LEVELS OF SCALE, GOOD SHAPE, POSITIVE SPACEfinding0.695Second cluster of the four-cluster grouping, containing properties 7, 13, 1, 6, 5.
- The degree to which centers are packed and overlapped, contributing to the life of the whole.
- Cluster 1 (4-cluster): CONTRAST, NOT-SEPARATENESS, ROUGHNESS, ALTERNATING REPETITION, GOOD SHAPEfinding0.694First cluster of the four-cluster grouping, containing properties 9, 15, 11, 4, 6.
- Approximate posterior probability distribution embodied in organism's internal states; organism's best guess about causes of sensations