A Mechanistic Interpretability Analysis of Grokking
This is a dump of relevant saved model weights and loss curves for the notebook A Mechanistic Interpretability Analysis of Grokking: https://colab.research.google.com/drive/1F6_1_cWXE5M7WocUcpQWp3v8z4b1jL20#scrollTo=rk8LtmzQfyDC&uniqifier=8