Downloads · 30 days
0
olmoe-sae-layer7
idimitrov/olmoe-sae-layer7
olmoe-sae-layer7 is a machine learning model from idimitrov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
This is a Sparse Autoencoder (SAE) trained on layer 7 of the allenai/OLMoE-1B-7B-0125 model.
Access
Public
Updated Oct 9, 2025
Repo size
537 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt134 MB · 50%
At a glance
- Access
- Public
- Created
- Oct 8, 2025
- Updated
- Oct 9, 2025
- SHA
- 998f8e06
Model card
From the Hugging Face model README
SAE for allenai/OLMoE-1B-7B-0125 - Layer 7
This is a Sparse Autoencoder (SAE) trained on layer 7 of the allenai/OLMoE-1B-7B-0125 model.
Model Details
- Base Model: allenai/OLMoE-1B-7B-0125
- Target Layer: 7
- Expansion Factor: 4x
- K-Sparsity: 32
Files
layers.7/sae.safetensorslayers.7/cfg.jsonconfig.jsonstate.ptoptimizer_0.ptrank_0_state.pt
- Created
- Oct 8, 2025
- Updated
- Oct 9, 2025