Downloads · 30 days
3
18% of all-time downloads
activation-oracle-multilayer-qwen3-4b-6L-3xlayer-loop-step-5000
nluick/activation-oracle-multilayer-qwen3-4b-6L-3xlayer-loop-step-5000
activation-oracle-multilayer-qwen3-4b-6L-3xlayer-loop-step-5000 is a machine learning model from nluick. 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.
This is a LoRA (Low-Rank Adaptation) adapter trained for SAE (Sparse Autoencoder) introspection tasks.
All-time downloads
17
Public
Repo size
540 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors529 MB · 97%
At a glance
- Library
- peft
- Model type
- qwen3
- Access
- Public
- Created
- Jan 5, 2026
- Updated
- Jan 5, 2026
- SHA
- 2835144d
Base models
Model card
From the Hugging Face model README
LoRA Adapter for SAE Introspection
This is a LoRA (Low-Rank Adaptation) adapter trained for SAE (Sparse Autoencoder) introspection tasks.
Base Model
- Base Model:
Qwen/Qwen3-4B - Adapter Type: LoRA
- Task: SAE Feature Introspection
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
# Load base model and tokenizer
base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B")
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-4B")
# Load LoRA adapter
model = PeftModel.from_pretrained(base_model, "nluick/activation-oracle-multilayer-qwen3-4b-6L-3xlayer-loop-step-5000")
Training Details
This adapter was trained using the lightweight SAE introspection training script to help the model understand and explain SAE features through activation steering.
- Library
- peft
- Type
- qwen3
- Created
- Jan 5, 2026
- Updated
- Jan 5, 2026