Mixture of Attention Heads: Selecting Attention Heads Per Token
31039Knowledge Triples
| Source Entity | Relationship | Target Entity |
|---|---|---|
|
Mixture of Attention Heads
Method |
evaluatedIn |
Machine Translation
Task |
|
Mixture of Attention Heads
Method |
evaluatedIn |
Masked Language Modeling
Method |
|
Mixture of Attention Heads
Method |
includes |
Attention Heads
Material |
|
Mixture of Attention Heads
Method |
includes |
Router
OtherEntity |