2026
Statistically Grounded Sparse-Feature Interventions for Activation-Space Control in Large Language Models
This work introduces a transparent SAE-feature steering pipeline that first applies a six-condition reliability filter, then ranks sparse features through an unweighted Borda consensus over three complementary statistics: F -test, KSG mutual information, and Cohen's d .
Oshayer Siddique, J. M. A. U. Alam, M. Rafy et al.
· arXiv.org · 0 citations