Update README.md

README.md

@@ -7,9 +7,9 @@ license: cc-by-nc-sa-4.0
 HypotheSAEs is a method which hypothesizes interpretable relationships in text datasets by training Sparse Autoencoders (SAEs) on foundation model representations.
 
 Paper: https://arxiv.org/abs/2502.04382  
-Code: https://github.com/rmovva/saetools-dev  
+Code: https://github.com/rmovva/HypotheSAEs
 
-This repo contains all of the datasets used for the experiments in the paper.
+This data repo contains all of the datasets used for the experiments in the paper.
 
 ## Overview