MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Tian, Ye, Yang, Ling, Yang, Jiongfan, Wang, Anran, Tian, Yu, Zheng, Jiani, Wang, Haochen, Teng, Zhiyang, Wang, Zhuochen, Wang, Yinjie, Tong, Yunhai, Wang, Mengdi, Li, Xiangtai
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918208028540928
author Tian, Ye
Yang, Ling
Yang, Jiongfan
Wang, Anran
Tian, Yu
Zheng, Jiani
Wang, Haochen
Teng, Zhiyang
Wang, Zhuochen
Wang, Yinjie
Tong, Yunhai
Wang, Mengdi
Li, Xiangtai
author_facet Tian, Ye
Yang, Ling
Yang, Jiongfan
Wang, Anran
Tian, Yu
Zheng, Jiani
Wang, Haochen
Teng, Zhiyang
Wang, Zhuochen
Wang, Yinjie
Tong, Yunhai
Wang, Mengdi
Li, Xiangtai
contents While thinking-aware generation aims to improve performance on complex tasks, we identify a critical failure mode where existing sequential, autoregressive approaches can paradoxically degrade performance due to error propagation. To systematically analyze this issue, we propose ParaBench, a new benchmark designed to evaluate both text and image output modalities. Our analysis using ParaBench reveals that this performance degradation is strongly correlated with poor alignment between the generated reasoning and the final image. To resolve this, we propose a parallel multimodal diffusion framework, MMaDA-Parallel, that enables continuous, bidirectional interaction between text and images throughout the entire denoising trajectory. MMaDA-Parallel is trained with supervised finetuning and then further optimized by Parallel Reinforcement Learning (ParaRL), a novel strategy that applies semantic rewards along the trajectory to enforce cross-modal consistency. Experiments validate that our model significantly improves cross-modal alignment and semantic consistency, achieving a 6.9\% improvement in Output Alignment on ParaBench compared to the state-of-the-art model, Bagel, establishing a more robust paradigm for thinking-aware image synthesis. Our code is open-sourced at https://github.com/tyfeld/MMaDA-Parallel
format Preprint
id arxiv_https___arxiv_org_abs_2511_09611
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation
Tian, Ye
Yang, Ling
Yang, Jiongfan
Wang, Anran
Tian, Yu
Zheng, Jiani
Wang, Haochen
Teng, Zhiyang
Wang, Zhuochen
Wang, Yinjie
Tong, Yunhai
Wang, Mengdi
Li, Xiangtai
Computer Vision and Pattern Recognition
While thinking-aware generation aims to improve performance on complex tasks, we identify a critical failure mode where existing sequential, autoregressive approaches can paradoxically degrade performance due to error propagation. To systematically analyze this issue, we propose ParaBench, a new benchmark designed to evaluate both text and image output modalities. Our analysis using ParaBench reveals that this performance degradation is strongly correlated with poor alignment between the generated reasoning and the final image. To resolve this, we propose a parallel multimodal diffusion framework, MMaDA-Parallel, that enables continuous, bidirectional interaction between text and images throughout the entire denoising trajectory. MMaDA-Parallel is trained with supervised finetuning and then further optimized by Parallel Reinforcement Learning (ParaRL), a novel strategy that applies semantic rewards along the trajectory to enforce cross-modal consistency. Experiments validate that our model significantly improves cross-modal alignment and semantic consistency, achieving a 6.9\% improvement in Output Alignment on ParaBench compared to the state-of-the-art model, Bagel, establishing a more robust paradigm for thinking-aware image synthesis. Our code is open-sourced at https://github.com/tyfeld/MMaDA-Parallel
title MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.09611