Hugging Face's logo

OpenMOSS-Team
/
Llama-Scope

English
Model card Files
xet
 Community
1
Hzfinfdu commited on
Commit
b39aad3
·
verified ·
1 Parent(s): 210b804

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -14,7 +14,7 @@ base_model:
14
 
15
  [**Use with SAELens** (In progress)]
16
 
17
- [**Explore in Neuronpedia** (In progress)]
18
 
19
  Sparse Autoencoders (SAEs) have emerged as a powerful unsupervised method for extracting sparse representations from language models, yet scalable training remains a significant challenge. We introduce a suite of 256 improved TopK SAEs, trained on each layer and sublayer of the Llama-3.1-8B-Base model, with 32K and 128K features.
20
 
 
14
 
15
  [**Use with SAELens** (In progress)]
16
 
17
+ [**Explore in Neuronpedia**](https://www.neuronpedia.org/llama-scope)
18
 
19
  Sparse Autoencoders (SAEs) have emerged as a powerful unsupervised method for extracting sparse representations from language models, yet scalable training remains a significant challenge. We introduce a suite of 256 improved TopK SAEs, trained on each layer and sublayer of the Llama-3.1-8B-Base model, with 32K and 128K features.
20