SyncAI.news, a Varaisys broadcasting
Don't Lose Focus: Activation Steering via Key-Orthogonal Projections
HL

Haoyan Luo, Mateo Espinosa Zarlenga, Mateja Jamnik

· 1 min read

ResearcharXiv cs.CL

Don't Lose Focus: Activation Steering via Key-Orthogonal Projections

arXiv:2605.06342v2 Announce Type: replace Abstract: Activation steering controls LLM behaviour towards target behaviour by intervening in internal representations, yet it often degrades reasoning and retrieval performance. We argue that a primary cause of this trade-off is attention rerouting: steering vectors alter query-key matching, shifting attention away from contextually important tokens toward less informative ones. To address this, we propose Steering via Key-Orthogonal Projections (SKOP), a steering method that constrains harmful attention rerouting without eliminating steering efficacy. SKOP achieves this by preserving attention patterns on a small set of focus tokens the model relies on for reasoning and retrieval, while allowing redistribution among less critical tail tokens. Across multiple steering benchmarks, we show that SKOP achieves the best joint steering-utility trade-off, reducing utility degradation by 5-7x while retaining over 95% of vanilla steering efficacy. Our results further suggest that, in long-context retrieval settings where vanilla steering approaches are ineffective, SKOP can maintain robust performance by avoiding attention rerouting.

Original source

This story was published by arXiv cs.CL and written by Haoyan Luo, Mateo Espinosa Zarlenga, Mateja Jamnik. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News