File size: 656 Bytes
a0d501e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
---
license: mit
library_name: pytorch
tags:
- coconut
- latent-chain-of-thought
- sparse-autoencoder
- qwen3
---

# SAE on Coconut latent thoughts (Qwen3-1.7B)

SAE bottleneck trained on the **frozen** Stage-3 Coconut latent thoughts of Qwen3-1.7B,
from the *VQ-CoT: Discretising Latent Chain-of-Thought* project (team RateLimit Achieved, EPFL CS-552).
The language model is frozen; only this bottleneck is trained on the dumped latents
(385,620 x 6 latent thought vectors, GSM8K).

- Checkpoint: `sae.pt`
- TopK SAE, dictionary 8x hidden, k=32 active features
- Result: inserting this bottleneck into the frozen latent loop costs ~0 GSM8K test accuracy.