SAELens
File size: 1,071 Bytes
604a214
 
5a81b10
604a214
 
5a81b10
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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 still uploading these SAEs

# 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-2b-pt-transcoders`?

- `gemma-scope-`: See 1.
- `2b-pt-`: These SAEs were trained on Gemma v2 2B base model.
- `transcoders`: These SAEs are transcoders: they were trained to reconstruct the output of MLP sublayers from the input to the MLP sublayers: see https://arxiv.org/abs/2406.11944

# 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