GitHub - pavanyellow/toy-counter-model: Interpretability of 2*2*2 layer model calculating the number of non zero entries in the input

pavanyellow / toy-counter-model Public

Interpretability of 2*2*2 layer model calculating the number of non zero entries in the input

Name		Name	Last commit message	Last commit date
Latest commit History 12 Commits
__pycache__		__pycache__
data		data
.DS_Store		.DS_Store
0.pt		0.pt
2.ipynb		2.ipynb
a.ipynb		a.ipynb
gpt2-sae.ipynb		gpt2-sae.ipynb
gpt2.ipynb		gpt2.ipynb
model.py		model.py
train.ipynb		train.ipynb
utils.py		utils.py

Provide feedback