finding
active
finding:ising-couplings-and-conditional-co-activation-give-cleanest-block-diagonal-affinity-for-colors-days-temperature-decoder-cosine-and-pearson-failIsing couplings and conditional co-activation give cleanest block-diagonal affinity for colors, days, temperature; decoder cosine and Pearson fail
Comparison of five similarity metrics.
Source paper
extracted_from(2026) · Usha Bhalla · Thomas Fel · Can Rager · Sheridan Feucht +8
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (2)
claim
- Why decoder cosine fails.
- Mere correlation of feature directions, as used in prior work, need not suffice to find manifoldssupportsContrast with Engels et al.
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.
- Coupling structure across sparsity regimes.
- Signature of each regime.
- Mechanistic evidence for two distinct representational pathways
- Post hoc recovery works synthetically.
- Practical finding: the method produces compact fixed-length representations from large volumes of token activations without requiring supervised labels.
- Prior finding from related work that aligns with ESR being strongest in the largest model tested
- Specific interpretive claim about what covariance pooling captures: the pairwise co-activation patterns across features that are invisible to mean pooling.