finding
active
finding:steering-induction-required-3-10-of-heads-and-fewer-than-half-as-many-when-non-steered-heads-attention-patterns-were-frozenSteering induction required 3–10% of heads, and fewer than half as many when non-steered heads' attention patterns were frozen
Evidence for hydra effect.
Source paper
extracted_fromNeighborhood — ranked by edge-count
Papers (1)
paper
Hypotheses (1)
hypothesis
- Explanation for steering counts.
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.
- The mechanistic explanation of how induction heads are implemented in two-layer models
- Central empirical claim of the paper; induction heads are shown to be the mechanism for powerful in-context learning
- Asymmetry suggesting more things must go right.
- 0% multi-attempt responses across 7,892 no-steering baseline trials confirming ESR is steering-inducedfinding0.762Control result establishing that self-correction is specifically induced by steering, not spontaneous model behavior
- Quantitative summary of Head Cor superiority on the Right-normalized Constrained Envelope Area metric
- Empirical comparison showing advantage of SAE features in low-data regime.
- Core applied contribution claim, supported by top-k accuracy comparisons.
- Practical disadvantage of activation steering highlighted as a drawback