A HyperNetwork is a network that generates weights for a main network. The behavior of the main network is the same 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 weight for that layer.
Source: HyperNetworksPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Federated Learning | 12 | 4.23% |
Multi-Task Learning | 10 | 3.52% |
Meta-Learning | 10 | 3.52% |
Decoder | 9 | 3.17% |
Personalized Federated Learning | 9 | 3.17% |
Continual Learning | 8 | 2.82% |
Few-Shot Learning | 8 | 2.82% |
Super-Resolution | 7 | 2.46% |
Reinforcement Learning (RL) | 7 | 2.46% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |