Diffusion Language Model Decoding

Follow this topic Ask about this topic

Momentum

26 papers in the last four weeks, up 271% on the four weeks before. 0.3% of all new papers.

Jul 13Week of Sep 28

Latest papers 164

All topics

DynaTE: Accelerating Diffusion LLMs via Dynamic Token Execution
Oct 8, 2026Minghan Jiang, Jiayi Wang, Shuaiting Li +2AI Accelerator Inference Diffusion Language Model Inference
College of Information Science & Electronic Engineering, Zhejiang University, Hangzhou 310027, China
Disentangling Paradigm, Identifier, and Decoding in Generative Retrieval
Oct 6, 2026Hicham Randrianarivo, Logan Renaud, Alexia AllalGenerative Retrieval Memorization in Generative Models
Artefact Research Center France
D-Loop: Looped Diffusion Drafting for Speculative Decoding
Oct 5, 2026Kecheng Chen, Yuyang He, Cheng Gong +8Speculative Decoding for Diffusion Language Models Diffusion Language Model Decoding
City University of Hong Kong · The Chinese University of Hong Kong · Huawei Research
Bayesian Entropy-based Reordering for Calibrated Diffusion Language Models
Oct 4, 2026Zhejun Jiang, Mijung ParkConfidence Estimation in Language Models Bayesian Inference
The University of British Columbia
Hierarchical Continuous Diffusion Language Models
Oct 1, 2026Hui Ren, Zihan Li, Chang Liu +2Diffusion Language Model Decoding Diffusion Language Models
University of Illinois Urbana-Champaign · Amazon.com, Inc.
Know When to Hold 'em: Correct-Token Retention in Uniform-State Diffusion Language Models
Oct 1, 2026Mojtaba Nafez, James HendersonDiffusion Language Model Decoding Diffusion Language Models
EPFL & Idiap Lausanne, Switzerland · Idiap Research Institute Martigny, Switzerland
Temporally-Resolved Token Attribution Reveals the Generation Dynamics of Diffusion Language Models
Oct 1, 2026Darpan Aswal, Céline HudelotIntegrated Gradients LLM Interpretability
Université Grenoble Alpes · MICS, CentraleSupélec
Fork-dLLM: Avoiding the Flexibility Trap in Diffusion Language Models
Sep 30, 2026Stipe Frković, Metod Jazbec, Christian A. NaessethMasked Diffusion Models Speculative Decoding for Diffusion Language Models
University of Amsterdam · UvA-Bosch Delta Lab, University of Amsterdam
Self-Repulsive Sampling for Diffusion Language Models
Sep 30, 2026Michael Helcig, Martin JaggiDiffusion Language Model Decoding Diverse Text Generation
ETH Zürich · EPFL
Does This Action Still Explain the Task? Reverse Scoring for Diffusion Language Model Agents
Sep 29, 2026Jiacheng Qiu, Christopher E. Mower, Jan Peters +2Diffusion Language Model Decoding LLM Agents
Huawei Noah’s Ark Lab · Technical University of Darmstadt · UCL Centre for AI
DScale: Scaling Block-Diffusion Speculative Decoding with Adaptive Verification
Sep 29, 2026Rongjian Chen, Minxian Xu, Zhengxin Fang +2Speculative Decoding LLM Inference Acceleration
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518052, China · University of Chinese Academy of Sciences, Beijing 100049, China · Victoria University of Wellington, Wellington, New Zealand +1
LongSpark: Efficient speculative decoding with a fixed-cost parallel drafter
Sep 29, 2026Hao-Yuan He, Peng-Fei Liu, Si Shen +1Long-Context Language Model Inference Blockwise Diffusion Language Models
Reliable Parallel Decoding in Masked Diffusion Language Models
Sep 29, 2026Zhenghao He, Bohan Liu, Guangzhi Xiong +1Masked Diffusion Models Diffusion Language Model Decoding
Department of Computer Science, University of Virginia
Twist, Don't Tilt: Trajectory-Exact Constrained Decoding for Masked Diffusion Models
Sep 28, 2026Aditya Thimmaiah, Lara Marinov, Jayanth Srinivasa +3Constrained Decoding Masked Diffusion Models
The University of Texas at Austin · Cisco Research
Why Deterministic PRM Guidance Underperforms in Discrete Diffusion Reasoning
Sep 28, 2026Yan Zhan, Shaobo Liu, Zhijun GaoProcess Reward Models Diffusion Language Model Decoding
Peking University · Shenzhen University
BV Loss: Block Verification-Aware Loss for Block Diffusion Speculative Decoding
Sep 28, 2026Suyoung Kim, Jahyun Koo, Hyeonjin Kim +5Speculative Decoding Diffusion Language Model Decoding
a2sys (a2sys.ai) · Seoul National University (snu.ac.kr) · University of Wisconsin (wisc.edu)
Low-Confidence Remasking Traps Flexibility: Realizing Arbitrary-Order Potential for Diverse Rollouts in Diffusion LLMs
Sep 28, 2026Moongyu Jeon, Dongjae Jeon, Bumjun Kim +2Masked Diffusion Models Diffusion Language Model Decoding
Yonsei University · KRAFTON AI · Kookmin University
From Position Risks to Block Survival: Faster Generation for Diffusion Language Models
Sep 27, 2026Siwei Chen, Yuxiang Wan, Yifan Yu +1Speculative Decoding for Diffusion Language Models Diffusion Language Model Decoding
University of Illinois Urbana-Champaign
Unmask the State: When Does State Adaptation Matter for Masked Diffusion Language Models
Sep 27, 2026Injin Kong, Sunghwan Choi, Yohan JoMasked Diffusion Models Diffusion Language Model Inference
Graduate School of Data Science, Seoul National University
Hesitation-Aware On-Policy Distillation for Diffusion Language Models
Sep 27, 2026Jianguo Huang, Lipeng Wan, Yanchen Deng +1Diffusion Language Model Decoding Uncertainty-Aware Knowledge Distillation
Nanyang Technological University, Singapore · Xi’an Jiaotong University, China
When Parallel Drafter Meets Parallel Speculative Decoding
Sep 23, 2026Fuliang Liu, Xue Li, Kun Qian +4Speculative Decoding LLM Inference Acceleration
State Key Laboratory of Novel Software Technology, Nanjing University · Alibaba Group
Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs
Sep 22, 2026Quan Nguyen-Tri, Mukul Ranjan, Zhiqiang ShenSpeculative Decoding Diffusion Language Model Inference
VILA Lab, MBZUAI Abu Dhabi, UAE
dQwen3.5: Hybrid-Attention Diffusion Language Models
Sep 17, 2026Anton Xue, Litu Rout, Aditya Akella +3Diffusion Language Model Decoding Diffusion Language Models
University of Texas at Austin
Parallelism, critical windows, and separations among diffusion language models
Sep 17, 2026Sitan Chen, Liye WangMasked Diffusion Models Diffusion Sampling
SEAS, Harvard University · Tsinghua University
Zarya: A Hybrid Autoregressive--Masked Diffusion Language Model with Flexible Training and Dual-Mode Inference
Sep 17, 2026Leonid Sinev, Ilya Koziev, Vladislav LeshchukMasked Diffusion Models Autoregressive Diffusion
Early-Bird Decoding: Accelerating Diffusion LLMs with Learnable Block Sizes and Parallel Sampling
Sep 15, 2026Lixuan Wei, Wei Zhou, Jianwen Wu +3Diffusion Language Model Inference Parallel Decoding
Harvard University · Georgia Institute of Technology · Purdue University
Temporal Self-Distillation: Faster Inference in Discrete Diffusion Language Models
Sep 14, 2026Shijian Xu, Andrea Miele, Metod Jazbec +3Diffusion Model Distillation Diffusion Language Model Inference
University of Basel · University of Amsterdam · Johns Hopkins University
DA-DLM: Explicitly Modeling Token Dependencies in Diffusion Language Models
Sep 14, 2026Pengyu Ji, Zichen Zhang, Xiang Hu +1Non-Autoregressive Text Generation Diffusion Language Model Decoding
‡School of Information Science and Technology, ShanghaiTech University · §Shanghai Engineering Research Center of Intelligent Vision and Imaging · ¶Tencent
Routing by Reasoning Need: Trajectory-Aware Decoding Control for Diffusion Vision-Language Models
Sep 12, 2026Yixiang Liu, Zhongxing Xu, Zhonghua Wang +1Diffusion Language Model Inference VLM Reasoning
Southern University of Science and Technology · Monash University
From Truncation to Commitment: Persistent Context in Uniform Discrete Diffusion
Sep 1, 2026Satoshi HayakawaDiffusion Sampling Diffusion Language Model Decoding
The University of Tokyo

Oct 8, 2026, cs.ARJ/K move · Enter open · S save

DynaTE: Accelerating Diffusion LLMs via Dynamic Token Execution

Minghan Jiang, Jiayi Wang, Shuaiting Li, Haibin Shen+1

College of Information Science & Electronic Engineering, Zhejiang University, Hangzhou 310027, China

AI Accelerator Inference Diffusion Language Model Inference Diffusion Language Model Decoding Hardware-Software Co-Design

Open paper AskPDF arXiv

Diffusion-based LLMs (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) LLMs by enabling bidirectional parallel refinement, alleviating the sequential decoding bottleneck of AR generation. However, their parallel iterative refinement mismatches AR accelerators optimized for sequential decoding and their discrete token generation differs from DiT accelerators designed for continuous denoising. Recent dLLM accelerators have explored workload-specific optimizations to reduce vocabulary processing overhead and redundant computation across denoising iterations. However, these approaches retain all tokens in parallel execution, despite varying token refinement utility and execution requirements. This paper presents DynaTE, a hardware--software co-design architecture that dynamically adapts accelerator execution to evolving token states during dLLM decoding. DynaTE first enables adaptive token execution by skipping low-utility token computation, while a dimension-reconfigurable PE array maintains high utilization under varying active-token patterns. Second, DynaTE exploits dynamic token dependencies through FLDD to refine a small number of locally dependent tokens within the current iteration, reducing the overall number of denoising iterations, while a Merge--Split--Merge dataflow hides the resulting serial overhead. Third, a streaming vocabulary engine interleaves multiple token streams from the LM head to accommodate irregular output variations caused by selective token computation and uneven vocabulary-selection demands. Evaluated on two representative dLLMs, DynaTE achieves 2.05--2.78

\times

speedup and 2.99--3.93

\times

higher energy efficiency over state-of-the-art dLLM accelerators, while delivering 2.55

\times

speedup and 6.07

\times

higher energy efficiency over Jetson AGX Orin.

Momentum

Related topics

Latest papers 164