finding
active
finding:distance-matrices-of-my-and-mz-correlate-r-71-p-001-for-arbitrary-domain-lower-similarityDistance matrices of My and Mz correlate r=.71 (p<.001) for Arbitrary domain, lower similarity
Weaker shared structure.
Source paper
extracted_from(2026) · Eric Bigelow · Raphaël Sarfati · Daniel Wurgaft · Owen Lewis +4
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (2)
claim
- Finding (2).
- Interpretation of Arbitrary results.
Hypotheses (1)
hypothesis
- Three concrete predictions about M.
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.
- Shared structure.
- Cross-architecture geometric invariance of Big Five steering vectors
- Appendix E replication of DIM alignment finding in Qwen model
- Core result of Experiment 3: cross-model semantic convergence under self-referential processing
- Key geometry-to-behavior bridge finding in E3; robust to pooling choice, cosine vs. L2, and frozen external encoder
- Country query vs own-city key cosine similarity mean −0.47, versus +0.01 for other citiesfinding0.748QK geometry anti-alignment.
- MAS reduces number of required alignment matrices for n-model comparison from n(n-1) or n^2 (stitching) to nfinding0.748Key computational efficiency advantage of MAS over traditional model stitching for multi-model comparisons.
- Fig. 9.