FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3100753857> ?p ?o ?g. }

• W3100753857 abstract "This paper presents two strong methods, CTC and Imputer, for non-autoregressive machine translation that model latent alignments with dynamic programming. We revisit CTC for machine translation and demonstrate that a simple CTC model can achieve state-of-the-art for single-step non-autoregressive machine translation, contrary to what prior work indicates. In addition, we adapt the Imputer model for non-autoregressive machine translation and demonstrate that Imputer with just 4 generation steps can match the performance of an autoregressive Transformer baseline. Our latent alignment models are simpler than many existing non-autoregressive translation baselines; for example, we do not require target length prediction or re-scoring with an autoregressive model. On the competitive WMT’14 En→De task, our CTC model achieves 25.7 BLEU with a single generation step, while Imputer achieves 27.5 BLEU with 2 generation steps, and 28.0 BLEU with 4 generation steps. This compares favourably to the autoregressive Transformer baseline at 27.8 BLEU." @default.
• W3100753857 created "2020-11-23" @default.
• W3100753857 creator A5000300454 @default.
• W3100753857 creator A5016918393 @default.
• W3100753857 creator A5018912210 @default.
• W3100753857 creator A5022455158 @default.
• W3100753857 date "2020-01-01" @default.
• W3100753857 modified "2023-09-30" @default.
• W3100753857 title "Non-Autoregressive Machine Translation with Latent Alignments" @default.
• W3100753857 cites W1574901103 @default.
• W3100753857 cites W2101105183 @default.
• W3100753857 cites W2127141656 @default.
• W3100753857 cites W2143612262 @default.
• W3100753857 cites W2148708890 @default.
• W3100753857 cites W2892213699 @default.
• W3100753857 cites W2920538220 @default.
• W3100753857 cites W2948629866 @default.
• W3100753857 cites W2952573656 @default.
• W3100753857 cites W2963250244 @default.
• W3100753857 cites W2963403868 @default.
• W3100753857 cites W2963434219 @default.
• W3100753857 cites W2963536265 @default.
• W3100753857 cites W2964121744 @default.
• W3100753857 cites W2971167892 @default.
• W3100753857 cites W2976965654 @default.
• W3100753857 cites W2982437993 @default.
• W3100753857 cites W2985671716 @default.
• W3100753857 cites W2988975212 @default.
• W3100753857 cites W2990372437 @default.
• W3100753857 cites W2990389671 @default.
• W3100753857 cites W2995999067 @default.
• W3100753857 cites W2999668723 @default.
• W3100753857 cites W3000538780 @default.
• W3100753857 cites W3000840023 @default.
• W3100753857 cites W3007746272 @default.
• W3100753857 cites W3015162217 @default.
• W3100753857 cites W3035289598 @default.
• W3100753857 cites W3035416964 @default.
• W3100753857 cites W3035490389 @default.
• W3100753857 doi "https://doi.org/10.18653/v1/2020.emnlp-main.83" @default.
• W3100753857 hasPublicationYear "2020" @default.
• W3100753857 type Work @default.
• W3100753857 sameAs 3100753857 @default.
• W3100753857 citedByCount "50" @default.
• W3100753857 countsByYear W31007538572020 @default.
• W3100753857 countsByYear W31007538572021 @default.
• W3100753857 countsByYear W31007538572022 @default.
• W3100753857 countsByYear W31007538572023 @default.
• W3100753857 crossrefType "proceedings-article" @default.
• W3100753857 hasAuthorship W3100753857A5000300454 @default.
• W3100753857 hasAuthorship W3100753857A5016918393 @default.
• W3100753857 hasAuthorship W3100753857A5018912210 @default.
• W3100753857 hasAuthorship W3100753857A5022455158 @default.
• W3100753857 hasBestOaLocation W31007538571 @default.
• W3100753857 hasConcept C104317684 @default.
• W3100753857 hasConcept C105580179 @default.
• W3100753857 hasConcept C105795698 @default.
• W3100753857 hasConcept C111368507 @default.
• W3100753857 hasConcept C119599485 @default.
• W3100753857 hasConcept C119857082 @default.
• W3100753857 hasConcept C12725497 @default.
• W3100753857 hasConcept C127313418 @default.
• W3100753857 hasConcept C127413603 @default.
• W3100753857 hasConcept C149364088 @default.
• W3100753857 hasConcept C151406439 @default.
• W3100753857 hasConcept C154945302 @default.
• W3100753857 hasConcept C159877910 @default.
• W3100753857 hasConcept C165801399 @default.
• W3100753857 hasConcept C185592680 @default.
• W3100753857 hasConcept C194657046 @default.
• W3100753857 hasConcept C203005215 @default.
• W3100753857 hasConcept C24338571 @default.
• W3100753857 hasConcept C28490314 @default.
• W3100753857 hasConcept C33923547 @default.
• W3100753857 hasConcept C41008148 @default.
• W3100753857 hasConcept C55493867 @default.
• W3100753857 hasConcept C622187 @default.
• W3100753857 hasConcept C66322947 @default.
• W3100753857 hasConceptScore W3100753857C104317684 @default.
• W3100753857 hasConceptScore W3100753857C105580179 @default.
• W3100753857 hasConceptScore W3100753857C105795698 @default.
• W3100753857 hasConceptScore W3100753857C111368507 @default.
• W3100753857 hasConceptScore W3100753857C119599485 @default.
• W3100753857 hasConceptScore W3100753857C119857082 @default.
• W3100753857 hasConceptScore W3100753857C12725497 @default.
• W3100753857 hasConceptScore W3100753857C127313418 @default.
• W3100753857 hasConceptScore W3100753857C127413603 @default.
• W3100753857 hasConceptScore W3100753857C149364088 @default.
• W3100753857 hasConceptScore W3100753857C151406439 @default.
• W3100753857 hasConceptScore W3100753857C154945302 @default.
• W3100753857 hasConceptScore W3100753857C159877910 @default.
• W3100753857 hasConceptScore W3100753857C165801399 @default.
• W3100753857 hasConceptScore W3100753857C185592680 @default.
• W3100753857 hasConceptScore W3100753857C194657046 @default.
• W3100753857 hasConceptScore W3100753857C203005215 @default.
• W3100753857 hasConceptScore W3100753857C24338571 @default.
• W3100753857 hasConceptScore W3100753857C28490314 @default.
• W3100753857 hasConceptScore W3100753857C33923547 @default.
• W3100753857 hasConceptScore W3100753857C41008148 @default.
• W3100753857 hasConceptScore W3100753857C55493867 @default.

• next