CircuitSteer builds cross-layer circuits from sparse autoencoder features using co-activation and decoder-direction alignment, then applies multi-layer steering vectors that it claims preserve fluency across all tested tasks.
Daniel Borkan, Lucas Dixon, Jeffrey Sorensen, Nithum Thain, and Lucy Vasserman
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
CircuitSteer: Geometrically Aligned Multi-Layer Steering via Sparse Autoencoder Circuits
CircuitSteer builds cross-layer circuits from sparse autoencoder features using co-activation and decoder-direction alignment, then applies multi-layer steering vectors that it claims preserve fluency across all tested tasks.