Improved Paraphrase Generation via Controllable Latent Diffusion

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zou, Wei, Zhuang, Ziyuan, Geng, Xiang, Huang, Shujian, Liu, Jia, Chen, Jiajun
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912192523141120
author Zou, Wei
Zhuang, Ziyuan
Geng, Xiang
Huang, Shujian
Liu, Jia
Chen, Jiajun
author_facet Zou, Wei
Zhuang, Ziyuan
Geng, Xiang
Huang, Shujian
Liu, Jia
Chen, Jiajun
contents Paraphrase generation strives to generate high-quality and diverse expressions of a given text, a domain where diffusion models excel. Though SOTA diffusion generation reconciles generation quality and diversity, textual diffusion suffers from a truncation issue that hinders efficiency and quality control. In this work, we propose \textit{L}atent \textit{D}iffusion \textit{P}araphraser~(LDP), a novel paraphrase generation by modeling a controllable diffusion process given a learned latent space. LDP achieves superior generation efficiency compared to its diffusion counterparts. It can facilitate only input segments to ensure paraphrase semantics, improving the results without external features. Experiments show that LDP better reconciles paraphrase generation quality and diversity than baselines. Further analysis shows that our method is also helpful to other similar text generations and domain adaptations
format Preprint
id arxiv_https___arxiv_org_abs_2404_08938
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Improved Paraphrase Generation via Controllable Latent Diffusion
Zou, Wei
Zhuang, Ziyuan
Geng, Xiang
Huang, Shujian
Liu, Jia
Chen, Jiajun
Computation and Language
Paraphrase generation strives to generate high-quality and diverse expressions of a given text, a domain where diffusion models excel. Though SOTA diffusion generation reconciles generation quality and diversity, textual diffusion suffers from a truncation issue that hinders efficiency and quality control. In this work, we propose \textit{L}atent \textit{D}iffusion \textit{P}araphraser~(LDP), a novel paraphrase generation by modeling a controllable diffusion process given a learned latent space. LDP achieves superior generation efficiency compared to its diffusion counterparts. It can facilitate only input segments to ensure paraphrase semantics, improving the results without external features. Experiments show that LDP better reconciles paraphrase generation quality and diversity than baselines. Further analysis shows that our method is also helpful to other similar text generations and domain adaptations
title Improved Paraphrase Generation via Controllable Latent Diffusion
topic Computation and Language
url https://arxiv.org/abs/2404.08938