Cover Image

Bringing Stability to Diffusion: Decomposing and Reducing Variance of Training Masked Diffusion Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Jia, Mengni, Zhou, Mengyu, Liu, Yihao, Jiang, Xiaoxi, Jiang, Guanjun
Format:	Preprint
Published:	2025
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Reconsidering Positional Supervision in Masked Diffusion Language Model Training
by: Ye, Mengyu, et al.
Published: (2026)

LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models
by: Zhu, Fengqi, et al.
Published: (2025)

Variance-Aware Adaptive Weighting for Diffusion Model Training
by: Sun, Nanlong, et al.
Published: (2026)

On the Trainability of Masked Diffusion Language Models via Blockwise Locality
by: Wang, Yuxiang, et al.
Published: (2026)

Training-Free Self-Correction for Multimodal Masked Diffusion Models
by: Ouyang, Yidong, et al.
Published: (2026)

Fill the GAP: A Granular Alignment Paradigm for Visual Reasoning in Multimodal Large Language Models
by: Miao, Yanting, et al.
Published: (2026)

MDPO: Overcoming the Training-Inference Divide of Masked Diffusion Language Models
by: He, Haoyu, et al.
Published: (2025)

LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models
by: Wei, Chenxing, et al.
Published: (2026)

DecompDiff: Diffusion Models with Decomposed Priors for Structure-Based Drug Design
by: Guan, Jiaqi, et al.
Published: (2024)

Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training
by: Kim, Jaeyeon, et al.
Published: (2026)

Optimizing Decoding Paths in Masked Diffusion Models by Quantifying Uncertainty
by: Chen, Ziyu, et al.
Published: (2025)

Fast Training of Diffusion Models with Masked Transformers
by: Zheng, Hongkai, et al.
Published: (2023)

Understanding and Accelerating the Training of Masked Diffusion Language Models
by: Hong, Chunsan, et al.
Published: (2026)

DecompOpt: Controllable and Decomposed Diffusion Models for Structure-based Molecular Optimization
by: Zhou, Xiangxin, et al.
Published: (2024)

Mask Is What DLLM Needs: A Masked Data Training Paradigm for Diffusion LLMs
by: Ma, Linrui, et al.
Published: (2026)

Auto-Regressive Masked Diffusion Models
by: Karami, Mahdi, et al.
Published: (2026)

Variance-Reduction Guidance: Sampling Trajectory Optimization for Diffusion Models
by: Xu, Shifeng, et al.
Published: (2025)

Train for the Worst, Plan for the Best: Understanding Token Ordering in Masked Diffusions
by: Kim, Jaeyeon, et al.
Published: (2025)

Beyond Masked and Unmasked: Discrete Diffusion Models via Partial Masking
by: Chao, Chen-Hao, et al.
Published: (2025)

Optimal Inference Schedules for Masked Diffusion Models
by: Chen, Sitan, et al.
Published: (2025)

Sharp Convergence Rates for Masked Diffusion Models
by: Liang, Yuchen, et al.
Published: (2026)

Stability and Generalization of Adversarial Diffusion Training
by: Hosseini, Hesam, et al.
Published: (2025)

Diffusion-State Policy Optimization for Masked Diffusion Language Models
by: Oba, Daisuke, et al.
Published: (2026)

Scaling Beyond Masked Diffusion Language Models
by: Sahoo, Subham Sekhar, et al.
Published: (2026)

DOS: Dependency-Oriented Sampler for Masked Diffusion Language Models
by: Zhou, Xueyu, et al.
Published: (2026)

Backdooring Masked Diffusion Language Models
by: Cao, Daniel Yiming, et al.
Published: (2026)

Masked Diffusion Modeling for Anomaly Detection
by: Zhang, Lixing, et al.
Published: (2026)

Self-Speculative Masked Diffusions
by: Campbell, Andrew, et al.
Published: (2025)

Variational Masked Diffusion Models
by: Zhang, Yichi, et al.
Published: (2025)

Disentangling Total-Variance and Signal-to-Noise-Ratio Improves Diffusion Models
by: Kahouli, Khaled, et al.
Published: (2025)

SliderSpace: Decomposing the Visual Capabilities of Diffusion Models
by: Gandikota, Rohit, et al.
Published: (2025)

No Compute Left Behind: Rethinking Reasoning and Sampling with Masked Diffusion Models
by: Horvitz, Zachary, et al.
Published: (2025)

A Decomposable Forward Process in Diffusion Models for Time-Series Forecasting
by: Caldas, Francisco, et al.
Published: (2026)

The Design Space of Tri-Modal Masked Diffusion Models
by: Bethune, Louis, et al.
Published: (2026)

BatteryBERT for Realistic Battery Fault Detection Using Point-Masked Signal Modeling
by: Zhou, Songqi, et al.
Published: (2025)

DualDiffusion: A Speculative Decoding Strategy for Masked Diffusion Models
by: Goyal, Satyam, et al.
Published: (2026)

Masked Diffusion Models are Secretly Learned-Order Autoregressive Models
by: Garg, Prateek, et al.
Published: (2025)

Revealing the Attention Floating Mechanism in Masked Diffusion Models
by: Dai, Xin, et al.
Published: (2026)

Eliminating Inductive Bias in Reward Models with Information-Theoretic Guidance
by: Li, Zhuo, et al.
Published: (2025)

On the Generalization Properties of Diffusion Models
by: Li, Puheng, et al.
Published: (2023)