Diffusion-based large language models (dLLMs) are gaining attention for their inherent capacity for parallel decoding, offering a compelling alternative to autoregressive LLMs. Among various decoding strategies, block-wise semi-autoregressive (semi-AR) approaches are widely adopted due to their support for KV caching and their favorable accuracy-speed trade-off. However, this paper identifies two fundamental limitations in the conventional semi-AR decoding approach that applies a fixed block size: i) late decoding overhead, where the unmasking of high-confidence tokens outside the current block is unnecessarily delayed, and ii) premature decoding error, where low-confidence tokens inside the current block are committed too early, leading to incorrect tokens. This paper presents the first systematic investigation challenging the fixed block size setting in semi-AR decoding. Through a statistical analysis of confidence dynamics during the denoising process, we identify a volatility band (VB) region during dLLM decoding, which encodes local semantic structure and can be used to guide adaptive block sizing. Leveraging these insights, we introduce AdaBlock-dLLM, a training-free, plug-and-play scheduler that adaptively aligns block boundaries with semantic steps by adjusting block size during runtime. Extensive experiments across diverse benchmarks show that AdaBlock-dLLM achieves up to 5.3% accuracy improvement under the same throughput budget. Beyond inference-time optimization, we hope our semantics-aware adaptive scheduling approach and confidence-based analysis will inspire future training strategies for dLLMs. Our code is available at https://github.com/lgxi24/AdaBlock-dLLM.
Streaming-dLLM: Accelerating Diffusion LLMs via Suffix Pruning and Dynamic Decoding
School of information（信息学院）；Electronics, Beijing Institute of Technology, Beijing（电子学院，北京理工大学，北京）；Department of Computer Science, City University of Hong Kong, Hong Kong（计算机科学系，香港城市大学，香港）；School of Computer Science, Wuhan University, Wuhan（计算机科学学院，武汉大学，武汉）；School of Economics（经济学院）；Management, Communication University of China, Beijing（管理学院，中国传媒大学，北京）；School of Science（科学学院）；Engineering, The Chinese University of Hong Kong (Shenzhen), Shenzhen（工程学院，香港中文大学（深圳），深圳）
SimSD: Simple Speculative Decoding in Diffusion Language Models
University of California San Diego（加州大学圣地亚哥分校）；University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Google（谷歌）；University of California Santa Barbara（加州大学圣芭芭拉分校）
d3LLM: Ultra-Fast Diffusion LLM using Pseudo-Trajectory Distillation
University of California, San Diego（加州大学圣地亚哥分校）；School of Artificial Intelligence, Nanjing University（南京大学人工智能学院）；National Key Laboratory for Novel Software Technology, Nanjing University（南京大学新型软件技术国家实验室）；Shanghai Jiao Tong University（上海交通大学）
LLaDA2.0: Scaling Up Diffusion Language Models to 100B
Ant Group（蚂蚁集团）；Renmin University of China（中国人民大学）；Zhejiang University（浙江大学）；Westlake University（西湖大学）；HongKong University of Science and Technology（香港科学与技术大学）
DiffuGuard: How Intrinsic Safety is Lost and Found in Diffusion Large Language Models
Beijing University of Posts and Telecommunications（北京邮电大学）；National University of Singapore（新加坡国立大学）；Nanyang Technological University（南洋理工大学）；Tsinghua University（清华大学）；The Chinese University of Hong Kong（香港中文大学）；Squirrel Ai Learning（松鼠AI）；Peking University（北京大学）
SLIM-RL: Risk-Budgeted Random-Masking RL for Diffusion LLMs Without Trajectory Slicing
Technical University of Denmark（丹麦技术大学）；MBZUAI Institute of Foundation Models（穆罕默德·本·扎耶德人工智能大学基础模型研究所）；Iowa State University（爱荷华州立大学）；Red Hat AI Innovation（红帽人工智能创新实验室）；MIT–IBM Watson AI Lab（麻省理工学院-IBM沃森人工智能实验室）
Escaping Confidence Trap: Evolutionary Decoding for Mathematical Reasoning in Diffusion LLMs
Australian National University（澳大利亚国立大学）；Nanyang Technological University（南洋理工大学）；National University of Singapore（新加坡国立大学）；Amazon（亚马逊）；University of Technology Sydney（悉尼科技大学）
STaR-Quant: State-Time Consistent Post-Training Quantization for Diffusion Large Language Models
School of Artificial Intelligence, Beijing Normal University, Beijing, China（北京师范大学人工智能学院）；Department of Computer Science, National University of Singapore, Singapore（新加坡国立大学计算机科学系）；Centre for Frontier AI Research, Agency for Science, Technology and Research (A*STAR), Singapore（科技研究局前沿人工智能研究中心）
GroupKV: Hierarchical KV Cache Management for Long-Context Diffusion LLM Inference
Institute of Software, Chinese Academy of Sciences（中国科学院软件研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Nanjing University of Science and Technology（南京理工大学）
Denoising Surface: Modeling and Predicting Inference Cost for Diffusion LLM Serving
Wuhan University（武汉大学）；Shanghai Jiao Tong University（上海交通大学）；The Hong Kong University of Science and Technology（香港科技大学）；Damen Database Co., Ltd.（达梦数据库有限公司）；Central China Normal University（华中师范大学）
Re-evaluating Confidence Remasking in Masked Diffusion Language Models
UvA-Bosch Delta Lab, University of Amsterdam（阿姆斯特丹大学UvA-Bosch Delta实验室）；Bosch Center for AI（博世人工智能中心）；University of Basel（巴塞尔大学）；Johns Hopkins University（约翰霍普金斯大学）
