Skip to content
Sagan

Paper

How Far Do Auto-Interpretation Labels Generalize: A Controlled Study Across Languages, Scripts, and Rewordings

Unreadunread

AI summary

arXiv:2606. 00356v1 Announce Type: new Abstract: Sparse autoencoder (SAE) features are increasingly used to interpret language models, with auto-generated natural-language labels serving as the primary interface for understanding what each feature represents.