UniReps
Blog
Toggle navigation
mechanistic interpretability
an archive of posts in this category
Feb 06, 2025
Do Sparse Autoencoders (SAEs) transfer across base and finetuned language models?