RSS

Amirali Abdullah

Karma: 5

Early Ex­per­i­ments in Re­ward Model In­ter­pre­ta­tion Us­ing Sparse Autoencoders

lukemarks, Amirali Abdullah, Rauno Arike, Fazl and nothoughtsheadempty
3 Oct 2023 7:45 UTC
11 points
0 comments5 min readLW link