Downloads · 30 days
0
analist/sae-r1-distill-qwen-1.5b-resid-l20-reasoning
sae-r1-distill-qwen-1.5b-resid-l20-reasoning is a machine learning model from analist. 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 sparsify. The card lists the license as mit.
A TopK sparse autoencoder trained from scratch on the residual stream (residpost) at layer 20 of deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B, over reasoning traces.
Downloads · 30 days
0
Access
Public
Updated Jul 13, 2026
Repo size
806 MB
Likes
0
Public
Click a slice to open those files.
.safetensors806 MB · 100%
From the Hugging Face model README
A TopK sparse autoencoder trained from scratch on the residual stream
(resid_post) at layer 20 of deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B,
over reasoning traces.
resid_post of decoder layer 20 (= output of model.model.layers[20],
TransformerLens blocks.20.hook_resid_post)from sparsify import Sae
sae = Sae.load_from_hub("analist/sae-r1-distill-qwen-1.5b-resid-l20-reasoning", hookpoint="resid_post_layer20", device="cuda")
Steering is intended at this layer (~71% relative depth), a region where reasoning-behavior features are steerable without producing token-level artifacts.
Trained with AuxK dead-latent resurrection and a unit-norm decoder; b_dec
initialized to the resid_post activation mean. See training_log.json for the
full config and FVU history.