finding
active
finding:inhibiting-dallas-features-reduces-texas-and-say-austin-features-but-leaves-say-a-capital-features-unaffectedInhibiting Dallas features reduces Texas and say-Austin features but leaves say-a-capital features unaffected.
Confirms graph-predicted effects.
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Central introductory claim.
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.
- If completion is mediated by a Texas step, swapping Texas features for another state should change the capital output.hypothesis0.766Prediction tested by feature swapping.
- Steering validation for antonym mechanism.
- Inhibiting RM bias features reduces 9-1-1 mentions to 6% and voting to 0%; baseline 42% 9-1-1, 56% voting.finding0.737Law-schools example.
- Perturbation validation.
- Meta-poem example.
- Author's mechanistic explanation for CV-CAA instability at higher injection strengths
- English is mechanistically the default output.
- §3.3.2.