Quantifying Feature Space Universality Across Large Language Models via Sparse Autoencoders

M. Lan, P. Torr, A. Meek, D. Krueger, F. Barez
October 2024
Type
Preprint
Publication
arXiv:2410.06981
Interpretability