Hugging Face
Search models, datasets, users...
Ctrl K
Models
Datasets
Spaces
Buckets
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
ceselder
's Collections
DeepSeek V4.1 Flash olens
Muse Spark olens
Kimi K3 olens
glm5.3-olens
Oracle lens (olens) + reconstructor + inverter, Qwen3.6-27B
Activation verbalizer + reconstructor pairs (NLA / AR)
Modulation Oracle
MAEMM cross-uplift matrix checkpoints (Qwen3.6-27B L42)
Qwen 3.6 27B good meta-models
LoRAcles: Weight-Space Interpretability at Scale
Skip-Lens: Multi-Token NLA Lenses (Qwen3.6-27B)
Building Better Activation Oracles
LoRAcle eval models
Building Better Activation Oracles
updated
Sep 1
Models and Datasets from Building Better Activation Oracles
Upvote
2
Sort: Collection
ceselder/qwen3-8b-ao-v3-adam-baseline
Text Generation
•
Updated
May 17
•
90
ceselder/qwen3-8b-ao-v3-abl1-sonnet-data
Text Generation
•
Updated
May 17
•
22
ceselder/qwen3-8b-ao-v3-abl2b-layer22
Text Generation
•
Updated
May 17
•
11
ceselder/qwen3-8b-ao-v3-abl2c-multilayer
Text Generation
•
Updated
May 17
•
19
ceselder/qwen3-8b-ao-v3-abl3-cotv5-corpus
Text Generation
•
Updated
May 17
•
22
ceselder/qwen3-8b-ao-v3-abl4c-sonnet-multilayer
Text Generation
•
Updated
May 17
•
16
ceselder/qwen3-8b-ao-v3-best
Text Generation
•
Updated
May 17
•
2.32k
•
1
ceselder/qwen3-8b-ao-v3-best-dpo-halluc
Text Generation
•
Updated
May 17
•
23
Upvote
2
Sort: Collection
Share collection
View history
Collection guide
Browse collections