SAELens
ArthurConmyGDM's picture
Update README.md
7ddcf83 verified
|
raw
history blame
1.11 kB
metadata
license: cc-by-4.0
library_name: saelens

⚠️ WARNING: We are in the process of uploading SAEs of many different sparsities for every (Layer, Width) pair. For now, there is only one sparsity per (Layer, Width) pair.

1. Gemma Scope

Gemma Scope is a comprehensive, open suite of sparse autoencoders for Gemma 2 9B and 2B. Sparse Autoencoders are a "microscope" of sorts that can help us break down a model’s internal activations into the underlying concepts, just as biologists use microscopes to study the individual cells of plants and animals.

See our landing page for details on the whole suite. This is a specific set of SAEs:

2. What Is gemma-scope-9b-pt-att?

  • gemma-scope-: See 1.
  • 9b-pt-: These SAEs were trained on Gemma v2 9B base model.
  • att: These SAEs were trained on the model's attention layer output before the linear projection.

3. Point of Contact

Point of contact: Arthur Conmy

Contact by email:

''.join(list('moc.elgoog@ymnoc')[::-1])

HuggingFace account: https://huggingface.co/ArthurConmyGDM