Login / Signup

Grokfast: Accelerated Grokking by Amplifying Slow Gradients.

Jaerin LeeBong Gyun KangKihoon KimKyoung Mu Lee
Published in: CoRR (2024)
Keyphrases
  • database
  • neural network
  • computer vision
  • multiscale
  • wide range
  • domain knowledge
  • gradient vector