Downloads · 30 days
0
SAE-Coconut-Qwen3-1.7B
Chengheng/SAE-Coconut-Qwen3-1.7B
SAE-Coconut-Qwen3-1.7B is a machine learning model from Chengheng. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for pytorch. The card lists the license as mit.
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; o…
Access
Public
Updated Jun 10, 2026
Repo size
269 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt269 MB · 100%
At a glance
- Library
- pytorch
- License
- mit
- Access
- Public
- Created
- Jun 10, 2026
- Updated
- Jun 10, 2026
- SHA
- a0d501e2
Model card
From the Hugging Face model README
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.
- Library
- pytorch
- License
- mit
- Created
- Jun 10, 2026
- Updated
- Jun 10, 2026