Downloads · 30 days
3
14% of all-time downloads
ceselder/olens-t15-n32-step30-qwen36-27b
olens-t15-n32-step30-qwen36-27b is a machine learning model from ceselder. 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 peft. The card lists the license as apache-2.0.
Oracle-lens LoRA for Qwen/Qwen3.6-27B: reads a raw layer-42 residual activation and emits four bullets of what the model is about to generate. 4-GPU DDP online best-of-N distillation from pi0, temp 1.5, N=32, step 30.…
Downloads · 30 days
3
14% of all-time downloads
All-time downloads
21
Public
Repo size
1.9 GB
Likes
0
Public
Click a slice to open those files.
.safetensors1.9 GB · 99%
From the Hugging Face model README
Oracle-lens LoRA for Qwen/Qwen3.6-27B: reads a raw layer-42 residual activation and emits four bullets of what the model is about to generate. 4-GPU DDP online best-of-N distillation from pi0, temp 1.5, N=32, step 30. Whitened FVE ~0.54 (single-activation info cap). Inject raw L42 at the marker token id 158983 (Karvonen norm-match); see nla_meta.yaml for the template.