README.md

---
base_model: google/gemma-2-9b-it
library_name: peft
---

# LoRA Adapter for SAE Introspection

This is a LoRA (Low-Rank Adaptation) adapter trained for SAE (Sparse Autoencoder) introspection tasks.

## Base Model
- **Base Model**: `google/gemma-2-9b-it`
- **Adapter Type**: LoRA
- **Task**: SAE Feature Introspection

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# Load base model and tokenizer
base_model = AutoModelForCausalLM.from_pretrained("google/gemma-2-9b-it")
tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-9b-it")

# Load LoRA adapter
model = PeftModel.from_pretrained(base_model, "thejaminator/gemma-hook-layer-0")
```

## 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.