##About

Publicly available artifacts from OpenAI's paper 
<strong>Weight-sparse transformers have interpretable circuits </strong>

https://cdn.openai.com/pdf/41df8f28-d4ef-43e9-aed2-823f9393e470/circuit-sparsity-paper.pdf