Login / Signup

HyperMoE: Paying Attention to Unselected Experts in Mixture of Experts via Dynamic Transfer.

Hao ZhaoZihan QiuHuijia WuZili WangZhaofeng HeJie Fu
Published in: CoRR (2024)
Keyphrases
  • paying attention
  • expert finding
  • mixture model
  • dynamically changing
  • real time
  • data sets
  • databases
  • neural network
  • video sequences