Hypernet起源于2017年 iclr的一篇文章 hypernetworks

In this work, we consider an approach of using a small network (called a “hypernetwork") to generate the weights for a larger network (called a main network). The behavior of the main network is the same as with any usual neural network: it learns to map some raw inputs to their desired targets; whereas the hypernetwork takes a set of inputs that contain information about the structure of the weights and generates the weights for that layer.

Hypernet 在 ​​Module Parameter Generation​​中起到了重要的作用



  1. ​https://aclanthology.org/2021.tacl-1.25/​
  2. ​https://proceedings.mlr.press/v162/he22f/he22f.pdf​
  3. ​https://aclanthology.org/2021.acl-long.47/​
  4. ​https://arxiv.org/abs/2205.12148​