Guardado en:
Detalles Bibliográficos
Autores principales: Tokarchuk, Evgeniia, Niculae, Vlad
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:https://arxiv.org/abs/2310.20620
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909156454170624
author Tokarchuk, Evgeniia
Niculae, Vlad
author_facet Tokarchuk, Evgeniia
Niculae, Vlad
contents Continuous-output neural machine translation (CoNMT) replaces the discrete next-word prediction problem with an embedding prediction. The semantic structure of the target embedding space (i.e., closeness of related words) is intuitively believed to be crucial. We challenge this assumption and show that completely random output embeddings can outperform laboriously pretrained ones, especially on larger datasets. Further investigation shows this surprising effect is strongest for rare words, due to the geometry of their embeddings. We shed further light on this finding by designing a mixed strategy that combines random and pre-trained embeddings for different tokens.
format Preprint
id arxiv_https___arxiv_org_abs_2310_20620
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle The Unreasonable Effectiveness of Random Target Embeddings for Continuous-Output Neural Machine Translation
Tokarchuk, Evgeniia
Niculae, Vlad
Computation and Language
Machine Learning
Continuous-output neural machine translation (CoNMT) replaces the discrete next-word prediction problem with an embedding prediction. The semantic structure of the target embedding space (i.e., closeness of related words) is intuitively believed to be crucial. We challenge this assumption and show that completely random output embeddings can outperform laboriously pretrained ones, especially on larger datasets. Further investigation shows this surprising effect is strongest for rare words, due to the geometry of their embeddings. We shed further light on this finding by designing a mixed strategy that combines random and pre-trained embeddings for different tokens.
title The Unreasonable Effectiveness of Random Target Embeddings for Continuous-Output Neural Machine Translation
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2310.20620