In recent years, the phenomenon of grokking—where deep learning models exhibit a delayed yet sudden transition from memorization to generalization—has prompted renewed investigation into training ...
This project is a hands-on exploration of the "grokking" phenomenon in neural networks, using a simple algorithmic task: modular addition (e.g., predicting (a + b) mod p). The project trains a small ...
According to God of Prompt (@godofprompt), grokking was first discovered by accident in 2022 when OpenAI researchers trained AI models on simple mathematical tasks such as modular addition and ...
According to God of Prompt on Twitter, DeepMind researchers have identified a phenomenon called 'Grokking' where neural networks may train for thousands of epochs with little to no improvement, then ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results