SAELens
File size: 1,138 Bytes
a93866d
be232c6
ebc655f
a93866d
 
d32eae0
 
be232c6
a93866d
be232c6
a93866d
be232c6
a93866d
be232c6
a93866d
be232c6
68cc5ce
be232c6
a93866d
be232c6
a93866d
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
---
license: cc-by-4.0
library_name: saelens
---

! WARNING: we are in the process of uploading these SAEs, and expect them to land in 30 mins. In the mean time the 2B SAEs, e.g. https://huggingface.co./google/gemma-scope-2b-pt SAEs should be available.

# 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](https://huggingface.co./google/gemma-scope) 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:

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

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