finding
active
finding:a-planning-feature-representing-end-word-rabbit-activates-at-the-newline-token-before-claude-begins-writing-a-rhyming-line-and-attention-streams-carry-it-forward-biasing-predictions-toward-rabbitA planning feature representing end-word 'rabbit' activates at the newline token before Claude begins writing a rhyming line, and attention streams carry it forward biasing predictions toward 'rabbit'
Illustrative finding from Lindsey et al. 2025 / Hanna and Ameisen 2026 showing intention-like representations in attention streams
Source paper
extracted_from(2026) · Pierre Beckmann · Patrick Butlin
Neighborhood — ranked by edge-count
Papers (1)
paper
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.
- Claude Opus 4.6 represents a plan to end a couplet with 'rabbit' before outputting the rhyming line.finding0.848Demonstrates causal relationship between NLA explanations and model outputs via steering with edited explanations.
- Quantitative planning intervention result.
- Tested via interventions at the newline planning site.
- Suppressing rabbit features changes the line to end with habit and alters intermediate transition.finding0.796Planned words shape sentence structure.
- Evidence that NLA explanations bear causal relationship to model outputs; demonstrates validity of extracted representations.
- Steering planned-word features only affects the final token when applied at the newline planning token.finding0.785Planning is localized.
- Planning in poems.
- Explanation for the 'silent' thought phenomenon.