Greedy Algorithm for Network Coarse-Graining

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Spectral Clustering for Network Coarse-Grainingmethod0.781
  Griebenow et al.'s method: eigenvalue decomposition of TPM, then OPTICS clustering to find macro-nodes.
• Coarse Grainingmethod0.768
• What coarse-graining strategy generates the maximum effective information?question0.764
  Central optimization problem in CE identification.
• Noisy Simulation of Sparse Networksconcept0.743
  Mechanism by which superposition works: small neural networks exploit sparsity to approximately simulate much larger sparse networks
• The feed-forward network truly implements a symbolic, tree-structured algorithm for hierarchical equality, with abstract equality relations not decomposable into input identitiesclaim0.731
  DAS reveals that the neural network encodes abstract relational structure rather than raw input identities.
• Evolutionary Algorithmsmethod0.721
  Machine learning approach using evolutionary processes to generate and select designs, used to blur the designed vs. evolved distinction
• Identification of algorithms implemented in attention layers, distributed across attention headsfinding0.715
  VPD successfully recovered interpretable attention algorithms (previous-token behavior, syntax-boundary routing) in weight space without requiring manual decomposition across heads.
• Neural Annealingframework0.713
  Michael Johnson's prior work on how neural networks (and brains) can be 'annealed' to find optimal states.