finding
active
finding:newline-feature-1254-tokens-features-and-ov-position-inputs-anti-correlated-ols-slope-0-993-r2-0-912-summing-to-roughly-constant-biasNewline feature 1254: Tokens→Features and OV position inputs anti-correlated (OLS slope -0.993, r²=0.912), summing to roughly constant bias
271 stratified contexts.
Source paper
extracted_fromNicholas L. Turner · Jeffrey Wu · Joshua Batson
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Circuit interpretation.
Questions (1)
question
- How does this feature know when to fire?answered_byNewline feature puzzle.
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.
- §5.5.
- Narrow-range greater-than feature.
- Turn-averaged features have highest embedding similarity 0.663 at k=3; per-token max lowest at 0.617finding0.763Consistent across k=1,2,3,5.
- User and Assistant emotions dissociate.
- Design choice validated empirically.
- Linear separability.
- LongBench v2 example E.
- Structured number feature.