Information Aggregation for Multi-Head Attention with Routing-by-Agreement

21932

Knowledge Triples

Source Entity	Relationship	Target Entity
Advanced Information Aggregation OtherEntity	comparesWith	Standard Linear Transformation OtherEntity
Linguistic Probing Tasks Task	evaluatedIn	Advanced Information Aggregation OtherEntity
Machine Translation Tasks Task	evaluatedIn	Advanced Information Aggregation OtherEntity
Multi-Head Attention Method	improves	Routing by Agreement Method
Routing by Agreement Method	proposes	Advanced Information Aggregation OtherEntity