method
active
method:hierarchical-clustering-with-dendrogramsHierarchical clustering with dendrograms
Used to test event alignment.
Neighborhood — ranked by edge-count
Papers (1)
paper
Frameworks (1)
framework
- Proposed objective decomposing activations into predictable and novel components.
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.
- Preliminary interpretability of interventions.
- Infers hierarchical structure from distance matrices.
- Preliminary mechanism evidence.
- Unsupervised feature-finding method using cluster centroid difference as feature direction
- Grouping similar model behaviors; the unsupervised method surfaces clusters of concerning patterns.
- Density-based clustering applied to 10-dimensional UMAP to organize feature directions into clusters
- Mode in feature density histogram around 1e-5 corresponding to interpretable features, contrasted with ultralow density cluster
- Griebenow et al.'s method: eigenvalue decomposition of TPM, then OPTICS clustering to find macro-nodes.