Static Word Embeddings for Sentence Semantic Representation

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Wada, Takashi, Hirakawa, Yuki, Shimizu, Ryotaro, Kawashima, Takahiro, Saito, Yuki
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909816183586816
author Wada, Takashi
Hirakawa, Yuki
Shimizu, Ryotaro
Kawashima, Takahiro
Saito, Yuki
author_facet Wada, Takashi
Hirakawa, Yuki
Shimizu, Ryotaro
Kawashima, Takahiro
Saito, Yuki
contents We propose new static word embeddings optimised for sentence semantic representation. We first extract word embeddings from a pre-trained Sentence Transformer, and improve them with sentence-level principal component analysis, followed by either knowledge distillation or contrastive learning. During inference, we represent sentences by simply averaging word embeddings, which requires little computational cost. We evaluate models on both monolingual and cross-lingual tasks and show that our model substantially outperforms existing static models on sentence semantic tasks, and even surpasses a basic Sentence Transformer model (SimCSE) on a text embedding benchmark. Lastly, we perform a variety of analyses and show that our method successfully removes word embedding components that are not highly relevant to sentence semantics, and adjusts the vector norms based on the influence of words on sentence semantics.
format Preprint
id arxiv_https___arxiv_org_abs_2506_04624
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Static Word Embeddings for Sentence Semantic Representation
Wada, Takashi
Hirakawa, Yuki
Shimizu, Ryotaro
Kawashima, Takahiro
Saito, Yuki
Computation and Language
Artificial Intelligence
Machine Learning
We propose new static word embeddings optimised for sentence semantic representation. We first extract word embeddings from a pre-trained Sentence Transformer, and improve them with sentence-level principal component analysis, followed by either knowledge distillation or contrastive learning. During inference, we represent sentences by simply averaging word embeddings, which requires little computational cost. We evaluate models on both monolingual and cross-lingual tasks and show that our model substantially outperforms existing static models on sentence semantic tasks, and even surpasses a basic Sentence Transformer model (SimCSE) on a text embedding benchmark. Lastly, we perform a variety of analyses and show that our method successfully removes word embedding components that are not highly relevant to sentence semantics, and adjusts the vector norms based on the influence of words on sentence semantics.
title Static Word Embeddings for Sentence Semantic Representation
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2506.04624