GitHub - pavanyellow/feature-steering: Controlling LLM outputs by activating/suppressing feature vectors

Name		Name	Last commit message	Last commit date
Latest commit History 30 Commits
.github		.github
.ipynb_checkpoints		.ipynb_checkpoints
__pycache__		__pycache__
data		data
frontend		frontend
LICENSE		LICENSE
README.md		README.md
a.py		a.py
activation-steering.ipynb		activation-steering.ipynb
fix_padding.ipynb		fix_padding.ipynb
generation.py		generation.py
get_steering_vector.ipynb		get_steering_vector.ipynb
get_steering_vector_honesty.ipynb		get_steering_vector_honesty.ipynb
get_steering_vector_yesman.ipynb		get_steering_vector_yesman.ipynb
index.html		index.html
llama-3-steering.ipynb		llama-3-steering.ipynb
manage_data.ipynb		manage_data.ipynb
model.py		model.py
requirements.txt		requirements.txt
run_llama-3.ipynb		run_llama-3.ipynb
setup.sh		setup.sh
tokenizer.py		tokenizer.py
use_steering_vector.ipynb		use_steering_vector.ipynb

Provide feedback