Neural Machine Translation in Linear Time

E899033

natural language processing paper research paper scientific publication

"Neural Machine Translation in Linear Time" is a research paper that introduces a more computationally efficient neural architecture for machine translation, reducing translation complexity to linear time with respect to input length.

Try in SPARQL Jump to: Surface forms Statements Referenced by

All labels observed (1)

Label	Occurrences
Neural Machine Translation in Linear Time canonical	1

Statements (33)

Predicate	Object
instanceOf	natural language processing paper ⓘ research paper ⓘ scientific publication ⓘ
addresses	scalability of neural machine translation to long input sequences ⓘ time complexity of neural machine translation models ⓘ
aimsTo	improve decoding speed in neural machine translation ⓘ maintain translation quality while reducing computational cost ⓘ reduce translation complexity from superlinear to linear in input length ⓘ
assumes	translation quality should not significantly degrade when reducing complexity ⓘ
comparesWith	more computationally expensive neural machine translation architectures ⓘ
concerns	decoding algorithms for neural machine translation ⓘ efficiency of translation models in practice ⓘ
contribution	analysis of complexity with respect to input length in neural translation ⓘ introduction of a linear-time neural architecture for sequence-to-sequence translation ⓘ
field	artificial intelligence ⓘ machine translation ⓘ natural language processing ⓘ neural machine translation ⓘ
focusesOn	computational efficiency in neural machine translation ⓘ reducing translation complexity to linear time with respect to input length ⓘ
optimizationGoal	faster inference for neural machine translation models ⓘ linear-time translation with respect to input length ⓘ reduced computational resources for translation ⓘ
proposes	a more computationally efficient neural architecture for machine translation ⓘ
relatedTo	attention mechanisms in neural networks ⓘ efficient neural architectures ⓘ scalable machine translation systems ⓘ sequence-to-sequence learning ⓘ time complexity analysis in neural models ⓘ
targets	longer input sentences ⓘ real-time or near real-time translation scenarios ⓘ
uses	neural networks ⓘ sequence-to-sequence modeling ⓘ

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.

Lukasz Kaiser → coAuthorOf → Neural Machine Translation in Linear Time ⓘ

subject surface form: Łukasz Kaiser