RSS

Georg Lange

Karma: 73

SAEs Dis­cover Mean­ingful Fea­tures in the IOI Task

Alex Makelov, Georg Lange and Neel Nanda
Jun 5, 2024, 11:48 PM
15 points
2 comments10 min readLW link

An In­ter­pretabil­ity Illu­sion for Ac­ti­va­tion Patch­ing of Ar­bi­trary Subspaces

Georg Lange, Alex Makelov and Neel Nanda
Aug 29, 2023, 1:04 AM
77 points
4 comments1 min readLW link