finding
active
finding:learning-of-6-features-in-3-dimensions-with-correlated-sets-proceeds-through-distinct-geometric-transformations-ending-in-octahedron-or-antiprismLearning of 6 features in 3 dimensions with correlated sets proceeds through distinct geometric transformations ending in octahedron or antiprism
Learning dynamics.
Neighborhood — ranked by edge-count
Papers (1)
paper
- Toy Models of Superpositionmentions
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.
- Both empirical and theoretical phase diagrams.
- Statistical grouping of properties based on dependency patterns, enabling deeper understanding of their coherence and interaction.
- Ensuing need for active, intelligent decoding on the expansion side.
- Decoder cosine similarity maps onto concept similarity.
- Abstract's closing thesis.
- Geometric configurations in uniform superposition.
- Fig. 9b.
- Empirical validation that PMI convergence actually occurs on real data