issues
search
ai-safety-foundation
/
sparse_autoencoder
Sparse Autoencoder for Mechanistic Interpretability
https://ai-safety-foundation.github.io/sparse_autoencoder/
MIT License
171
stars
39
forks
source link
Ensure that the adam reset indices are on the correct device
#174
Closed
alan-cooney
closed
8 months ago