A Mechanistic Interpretability Analysis of Grokking
alignmentforum.org
A Mechanistic Interpretability Analysis of Grokking
1–2 of 2 posts
Re: A Mechanistic Interpretability Analysis of Grokking
#2related:
Grokking: Generalization beyond overfitting on small algorithmic datasets https://mathai-iclr.github.io/papers/papers/MATHAI_29_paper....