Saved in:
Bibliographic Details
Main Authors: Yu, Chuanyue, Wang, Jiahui, Li, Yuhan, Chang, Heng, Lan, Ge, Sun, Qingyun, Li, Jia, Li, Jianxin, Zhang, Ziwei
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.11342
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909992539389952
author Yu, Chuanyue
Wang, Jiahui
Li, Yuhan
Chang, Heng
Lan, Ge
Sun, Qingyun
Li, Jia
Li, Jianxin
Zhang, Ziwei
author_facet Yu, Chuanyue
Wang, Jiahui
Li, Yuhan
Chang, Heng
Lan, Ge
Sun, Qingyun
Li, Jia
Li, Jianxin
Zhang, Ziwei
contents Diffusion Language Models (DLMs) have recently demonstrated remarkable capabilities in natural language processing tasks. However, the potential of Retrieval-Augmented Generation (RAG), which shows great successes for enhancing large language models (LLMs), has not been well explored, due to the fundamental difference between LLM and DLM decoding. To fill this critical gap, we systematically test the performance of DLMs within the RAG framework. Our findings reveal that DLMs coupled with RAG show promising potentials with stronger dependency on contextual information, but suffer from limited generation precision. We identify a key underlying issue: Response Semantic Drift (RSD), where the generated answer progressively deviates from the query's original semantics, leading to low precision content. We trace this problem to the denoising strategies in DLMs, which fail to maintain semantic alignment with the query throughout the iterative denoising process. To address this, we propose Semantic-Preserving REtrieval-Augmented Diffusion (SPREAD), a novel framework that introduces a query-relevance-guided denoising strategy. By actively guiding the denoising trajectory, SPREAD ensures the generation remains anchored to the query's semantics and effectively suppresses drift. Experimental results demonstrate that SPREAD significantly enhances the precision and effectively mitigates RSD of generated answers within the RAG framework.
format Preprint
id arxiv_https___arxiv_org_abs_2601_11342
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Unlocking the Potentials of Retrieval-Augmented Generation for Diffusion Language Models
Yu, Chuanyue
Wang, Jiahui
Li, Yuhan
Chang, Heng
Lan, Ge
Sun, Qingyun
Li, Jia
Li, Jianxin
Zhang, Ziwei
Machine Learning
Computation and Language
Diffusion Language Models (DLMs) have recently demonstrated remarkable capabilities in natural language processing tasks. However, the potential of Retrieval-Augmented Generation (RAG), which shows great successes for enhancing large language models (LLMs), has not been well explored, due to the fundamental difference between LLM and DLM decoding. To fill this critical gap, we systematically test the performance of DLMs within the RAG framework. Our findings reveal that DLMs coupled with RAG show promising potentials with stronger dependency on contextual information, but suffer from limited generation precision. We identify a key underlying issue: Response Semantic Drift (RSD), where the generated answer progressively deviates from the query's original semantics, leading to low precision content. We trace this problem to the denoising strategies in DLMs, which fail to maintain semantic alignment with the query throughout the iterative denoising process. To address this, we propose Semantic-Preserving REtrieval-Augmented Diffusion (SPREAD), a novel framework that introduces a query-relevance-guided denoising strategy. By actively guiding the denoising trajectory, SPREAD ensures the generation remains anchored to the query's semantics and effectively suppresses drift. Experimental results demonstrate that SPREAD significantly enhances the precision and effectively mitigates RSD of generated answers within the RAG framework.
title Unlocking the Potentials of Retrieval-Augmented Generation for Diffusion Language Models
topic Machine Learning
Computation and Language
url https://arxiv.org/abs/2601.11342