Edit k-NN

Computes edit distance required to match nearest neighbors between two datasets, normalized by maximum edit distance

Neighborhood — ranked by edge-count

method

Edit Distance k-NN
related_to
Alternative alignment metric compared in appendix; computes edit distance between nearest neighbor lists

cosine ≥ 0.65 · no typed edge

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

Cycle k-NNmethod0.789
Alternative alignment metric; measures whether nearest neighbor in one domain also considers original sample as nearest neighbor in other domain
LCS k-NNmethod0.780
Calculates the longest common subsequence of nearest neighbors normalized by sequence length
Longest Common Subsequence k-NNmethod0.763
Alternative alignment metric compared in appendix; calculates longest common subsequence of nearest neighbor lists
K-Compositionconcept0.710
A form of attention head composition where W_K reads from a subspace affected by a previous head; central to how induction heads are implemented
Model Editingconcept0.709
Technique for modifying model knowledge or behavior via targeted interventions, e.g., ROME by Meng et al.
Manual model editing through parameter manipulationconcept0.691
Application enabled by VPD: direct manipulation of weight matrices for interpretable model modification.
N-gramsmethod0.687
Statistical model of next-letter probabilities used by Shannon.
n-way Associative Lookupconcept0.682
The single primitive operation of looking up a value from a set of n keys.