Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
SAE-MInAlA_Llama-3.2-3B-SimPO-merged-layer_14_mid – AI Model by chungimungi | AlphaNeural AI
You can deploy this model and start earning money today!
chungimungi
/
SAE-MInAlA_Llama-3.2-3B-SimPO-merged-layer_14_mid
like
0
pytorch
sae
sparse-autoencoder
interpretability
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
chungimungi/SAE-MInAlA_Llama-3.2-3B-SimPO-merged-layer_14_mid
Batch Top-K SAE trained on
MInAlA/Llama-3.2-3B-SimPO-merged
at
model.layers.14
(mid layer), matching the interp-alignment SAE training config (d_sae=16384, k=64, 2M tokens, UltraChat).
