Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
SKwra
/
toolcalling-sae
like
0
sparse-autoencoder
mechanistic-interpretability
tool-calling
gemma
ministral
qwen
arxiv:
2605.18882
License:
apache-2.0
Model card
Files
Files and versions
xet
Community
Copy to bucket
new
main
toolcalling-sae
/
Qwen3.5-9B
/
stage2
537 MB
Ctrl+K
Ctrl+K
1 contributor
History:
1 commit
SKwra
Add SAE checkpoints (all models, stage1 + stage2)
1cdd4de
verified
5 days ago
Qwen3.5-9B-L25-d32768-5M-stage2.pt
pickle
Detected Pickle imports (4)
"torch.BFloat16Storage"
,
"sae.sae_model.SAEConfig"
,
"collections.OrderedDict"
,
"torch._utils._rebuild_tensor_v2"
How to fix it?
537 MB
xet
Add SAE checkpoints (all models, stage1 + stage2)
5 days ago
Qwen3___5-9B-L25-d32768-5M-stage2_stats.json
3.48 kB
Add SAE checkpoints (all models, stage1 + stage2)
5 days ago
stage2_data_preview.txt
11.7 kB
Add SAE checkpoints (all models, stage1 + stage2)
5 days ago
