Fuse It More Deeply! A Variational Transformer with Layer-Wise Latent Variable Inference for Text Generation
21455Knowledge Triples
| Source Entity | Relationship | Target Entity |
|---|---|---|
|
Auto Regressive Decoders
Method |
contributesTo |
KL Vanishing Problem
Task |
|
Della
Method |
improves |
Generation Tasks
Task |
|
Della
Method |
isBasedOn |
Transformer
Method |
|
Della
Method |
uses |
Latent Variables
Material |
|
KL Vanishing Problem
Task |
addressedBy |
Della
Method |
|
KL Vanishing Problem
Task |
measuredBy |
KL Values
Metric |