---
title: "CreditDecoding — paper by Wang et al. (2026)"
description: "They define Trace Credit, a per-token score built by accumulating logits from previous steps. 33 references."
authors: ["Kangyu Wang", "Zhiyun Jiang", "Haibo Feng", "Weijia Zhao", "Lin Liu", "Jianguo Li", "Zhenzhong Lan", "Weiyao Lin"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.509/"
doi: "10.18653/v1/2026.acl-long.509"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/wang-2026-creditdecoding-accelerating-parallel-decoding-diffusion/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/wang-2026-creditdecoding-accelerating-parallel-decoding-diffusion/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/26/) / CreditDecoding: Accelerating Parallel Decoding in Diffusion Large Language Models with Trace Credit

# CreditDecoding: Accelerating Parallel Decoding in Diffusion Large Language Models with Trace Credit

**Authors:** Kangyu Wang, Zhiyun Jiang, Haibo Feng, Weijia Zhao, [Lin Liu](https://md.research.elamur.ai/free/en/authors/lin-liu/), Jianguo Li, [Zhenzhong Lan](https://md.research.elamur.ai/free/en/authors/zhenzhong-lan/) and Weiyao Lin · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.acl-long.509/) · [PDF](https://aclanthology.org/2026.acl-long.509.pdf) · [DOI 10.18653/v1/2026.acl-long.509](https://doi.org/10.18653/v1/2026.acl-long.509) · [OpenAlex](https://openalex.org/W4414981911)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Diffusion large language models (dLLMs) generate text by repeatedly denoising a masked sequence, but inference is slow. The usual threshold-based parallel decoding commits only tokens whose current confidence passes a cutoff and remasks the rest. By studying denoising traces, the authors find that the final output token is often already the model's top guess several steps before its confidence is high enough to be accepted. As a result, correct tokens are remasked and re-predicted many times, and each step ignores earlier predictions.

They define Trace Credit, a per-token score built by accumulating logits from previous steps. On top of it they build CreditDecoding, a training-free method that adds this credit to the current logits as a prior, so tokens that are predicted consistently but with low confidence get decoded sooner.

Across four dLLMs and eight benchmarks covering knowledge, reasoning and coding, the method reaches up to 5.48× speedup on LLaDA-8B-Instruct with a small accuracy gain (+0.48). Because it only post-processes logits, it can be added to existing optimizations, which makes it a practical inference add-on.

## Key points
- Key observation: target tokens become top-1 predictions well before they are decoded. In one illustrated case (position 128), adding credit shortens decoding by about 6.2 steps.
- On LLaDA-8B-Instruct, CreditDecoding gives up to 5.48× speedup and +0.48 average score. The authors also propose a tuning-free variant for easier use.
- It works alongside threshold decoding (Fast-dLLM), KV cache variants, kernel/compiler-level acceleration and EOS early stopping. Combined with EOS early stopping, it reaches 4.7× acceleration over the baseline with +0.63 average score.
- In the decoding-rule comparison (Table 3):
  - Threshold 0.95: score 60.38 → 60.86, tokens per forward pass (TPF) 5.49 → 7.00.
  - Top-K margin: score 60.44 → 61.20, TPF 5.49 → 6.27.
  - Fixed Top-K (K = 2, 4, 8): speed cannot change, but scores still rise slightly.
- In long-context tests on LLaDA (generation lengths up to 4k), it keeps a roughly 7× TPF speedup. It also degrades less than the baseline as generation length grows.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [🔌 Hardware and Architecture](https://md.research.elamur.ai/free/en/shelves/computer-science/hardware-and-architecture/) · 81 works

🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works

## Cite

Cite as: Kangyu Wang, Zhiyun Jiang, Haibo Feng, Weijia Zhao, Lin Liu, Jianguo Li, Zhenzhong Lan and Weiyao Lin. 2026. CreditDecoding: Accelerating Parallel Decoding in Diffusion Large Language Models with Trace Credit. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{wang2026creditdecoding,
  title = {{CreditDecoding: Accelerating Parallel Decoding in Diffusion Large Language Models with Trace Credit}},
  author = {Kangyu Wang, Zhiyun Jiang, Haibo Feng, Weijia Zhao, Lin Liu, Jianguo Li, Zhenzhong Lan and Weiyao Lin},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.509},
  url = {https://doi.org/10.18653/v1/2026.acl-long.509}
}
```

- [← previous Credibility Assessment for Arabic News on the Gaza War: A Hybrid Neural-Symbolic Pipeline](https://md.research.elamur.ai/free/en/papers/abril-2026-credibility-assessment-arabic-news-gaza/)
- [next → CRiT-QA: Evaluating Multi-hop Reasoning with Counterfactual Chains and Distractor Traps](https://md.research.elamur.ai/free/en/papers/yun-2026-crit-qa-evaluating-multi-hop/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/26/)

## References

33 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Tiwei Bie, Maosong Cao, Kun Chen, Lun Du, Mingliang Gong, Zhuochen Gong, Yanmei Gu, Jiaqi Hu, Zenan Huang, Zhenzhong Lan, Chengxi Li, Chongxuan Li, Jianguo Li, Zehuan Li, Huabin Liu, Lin Liu, Guoshan Lu, Xiaocheng Lu, Yuxin Ma, and 12 others. 2025. Llada2.0: Scaling up diffusion language models to 1… — [arXiv:2512.15745](https://arxiv.org/abs/2512.15745) [find](https://scholar.google.com/scholar?q=Tiwei+Bie%2C+Maosong+Cao%2C+Kun+Chen%2C+Lun+Du%2C+Mingliang+Gong%2C+Zhuochen+Gong%2C+Yanmei+Gu%2C+Jiaqi+Hu%2C+Zenan+Huang%2C+Zhenzhong+Lan%2C+Chengxi+Li%2C+Chongxuan+Li%2C+Jianguo+Li%2C+Zehuan+Li%2C+Huabin+Liu%2C+Lin+Liu%2C+Guoshan+)
2. Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde de Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, Alex Ray, Raul Puri, Gretchen Krueger, Michael Petrov, Heidy Khlaaf, Girish Sastry, Pamela Mishkin, Brooke Chan, Scott Gray, and 39 others. 202… — [Evaluating Large Language Models Trained on Code](https://md.research.elamur.ai/free/en/papers/chen-2021-evaluating-large-language-models-trained/)
3. Xinhua Chen, Sitao Huang, Cong Guo, Chiyue Wei, Yintao He, Jianyi Zhang, Hai "Helen" Li, and Yiran Chen. 2025. Dpad: Efficient diffusion language models with suffix dropout. Preprint, arXiv:2508.14148. — [arXiv:2508.14148](https://arxiv.org/abs/2508.14148) [find](https://scholar.google.com/scholar?q=Xinhua+Chen%2C+Sitao+Huang%2C+Cong+Guo%2C+Chiyue+Wei%2C+Yintao+He%2C+Jianyi+Zhang%2C+Hai+%22Helen%22+Li%2C+and+Yiran+Chen%2E+2025%2E+Dpad%3A+Efficient+diffusion+language+models+with+suffix+dropout%2E+Preprint%2C+arXiv%3A2508%2E14148)
4. Shuang Cheng, Yihan Bian, Dawei Liu, Linfeng Zhang, Qian Yao, Zhongbo Tian, Wenhai Wang, Qipeng Guo, Kai Chen, Biqing Qi, and Bowen Zhou. 2025. Sdar: A synergistic diffusion-autoregression paradigm for scalable sequence generation. Preprint, arXiv:2510.06303. [find](https://scholar.google.com/scholar?q=Shuang+Cheng%2C+Yihan+Bian%2C+Dawei+Liu%2C+Linfeng+Zhang%2C+Qian+Yao%2C+Zhongbo+Tian%2C+Wenhai+Wang%2C+Qipeng+Guo%2C+Kai+Chen%2C+Biqing+Qi%2C+and+Bowen+Zhou%2E+2025%2E+Sdar%3A+A+synergistic+diffusion%2Dautoregression+paradigm+fo)
5. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. 2021. [Training verifiers to solve math word problems](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). Preprint, arXiv:2110.14168.
6. Dheeru Dua, Yizhong Wang, Pradeep Dasigi, Gabriel Stanovsky, Sameer Singh, and Matt Gardner. 2019. [Drop: A reading comprehension benchmark requiring discrete reasoning over paragraphs](https://md.research.elamur.ai/free/en/papers/dua-2019-drop-reading-comprehension-benchmark-requiring/). In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistic…
7. Guhao Feng, Yihan Geng, Jian Guan, Wei Wu, Liwei Wang, and Di He. 2025. Theoretical benefit and limitation of diffusion language model. Preprint, arXiv:2502.09622. — [arXiv:2502.09622](https://arxiv.org/abs/2502.09622) [find](https://scholar.google.com/scholar?q=Guhao+Feng%2C+Yihan+Geng%2C+Jian+Guan%2C+Wei+Wu%2C+Liwei+Wang%2C+and+Di+He%2E+2025%2E+Theoretical+benefit+and+limitation+of+diffusion+language+model%2E+Preprint%2C+arXiv%3A2502%2E09622%2E+%E2%80%94+%5BarXiv%3A2502%2E09622%5D%28https%3A%2F%2Farxiv%2Eo)
8. Shansan Gong, Shivam Agarwal, Yizhe Zhang, Jiacheng Ye, Lin Zheng, Mukai Li, Chenxin An, Peilin Zhao, Wei Bi, Jiawei Han, Hao Peng, and Lingpeng Kong. 2025a. Scaling diffusion language models via adaptation from autoregressive models. Preprint, arXiv:2410.17891. — [arXiv:2410.17891](https://arxiv.org/abs/2410.17891) [find](https://scholar.google.com/scholar?q=Shansan+Gong%2C+Shivam+Agarwal%2C+Yizhe+Zhang%2C+Jiacheng+Ye%2C+Lin+Zheng%2C+Mukai+Li%2C+Chenxin+An%2C+Peilin+Zhao%2C+Wei+Bi%2C+Jiawei+Han%2C+Hao+Peng%2C+and+Lingpeng+Kong%2E+2025a%2E+Scaling+diffusion+language+models+via+adap)
9. Shansan Gong, Ruixiang Zhang, Huangjie Zheng, Jiatao Gu, Navdeep Jaitly, Lingpeng Kong, and Yizhe Zhang. 2025b. Diffucoder: Understanding and improving masked diffusion models for code generation. Preprint, arXiv:2506.20639. — [arXiv:2506.20639](https://arxiv.org/abs/2506.20639) [find](https://scholar.google.com/scholar?q=Shansan+Gong%2C+Ruixiang+Zhang%2C+Huangjie+Zheng%2C+Jiatao+Gu%2C+Navdeep+Jaitly%2C+Lingpeng+Kong%2C+and+Yizhe+Zhang%2E+2025b%2E+Diffucoder%3A+Understanding+and+improving+masked+diffusion+models+for+code+generation%2E+Pre)
10. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021a. [Measuring massive multitask language understanding](https://md.research.elamur.ai/free/en/papers/hendrycks-2020-measuring-massive-multitask-language-understanding/). Preprint, arXiv:2009.03300.
11. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021b. [Measuring mathematical problem solving with the math dataset](https://md.research.elamur.ai/free/en/papers/hendrycks-2021-measuring-mathematical-problem-solving-math/). Preprint, arXiv:2103.03874.
12. Pengcheng Huang, Shuhao Liu, Zhenghao Liu, Yukun Yan, Shuo Wang, Zulong Chen, and Tong Xiao. 2025. Pc-sampler: Position-aware calibration of decoding bias in masked diffusion models. Preprint, arXiv:2508.13021. — [arXiv:2508.13021](https://arxiv.org/abs/2508.13021) [find](https://scholar.google.com/scholar?q=Pengcheng+Huang%2C+Shuhao+Liu%2C+Zhenghao+Liu%2C+Yukun+Yan%2C+Shuo+Wang%2C+Zulong+Chen%2C+and+Tong+Xiao%2E+2025%2E+Pc%2Dsampler%3A+Position%2Daware+calibration+of+decoding+bias+in+masked+diffusion+models%2E+Preprint%2C+arXiv%3A2)
13. Animesh Jain, Shunting Zhang, Edward Yang, and 1 others. 2023. Pytorch 2.0: Our next generation 2.0 release. [find](https://scholar.google.com/scholar?q=Animesh+Jain%2C+Shunting+Zhang%2C+Edward+Yang%2C+and+1+others%2E+2023%2E+Pytorch+2%2E0%3A+Our+next+generation+2%2E0+release%2E)
14. Yuxin Ma, Lun Du, Lanning Wei, Kun Chen, Qian Xu, Kangyu Wang, Guofeng Feng, Guoshan Lu, Lin Liu, Xiaojing Qi, Xinyuan Zhang, Zhen Tao, Haibo Feng, Ziyun Jiang, Ying Xu, Zenan Huang, Yihong Zhuang, Haokai Xu, Jiaqi Hu, and 4 others. 2025b. dinfer: An efficient inference framework for diffusion langu… [find](https://scholar.google.com/scholar?q=Yuxin+Ma%2C+Lun+Du%2C+Lanning+Wei%2C+Kun+Chen%2C+Qian+Xu%2C+Kangyu+Wang%2C+Guofeng+Feng%2C+Guoshan+Lu%2C+Lin+Liu%2C+Xiaojing+Qi%2C+Xinyuan+Zhang%2C+Zhen+Tao%2C+Haibo+Feng%2C+Ziyun+Jiang%2C+Ying+Xu%2C+Zenan+Huang%2C+Yihong+Zhuang%2C+Ha)
15. Shen Nie, Fengqi Zhu, Zebin You, Xiaolu Zhang, Jingyang Ou, Jun Hu, Jun Zhou, Yankai Lin, Ji-Rong Wen, and Chongxuan Li. 2025. [Large language diffusion models](https://md.research.elamur.ai/free/en/papers/nie-2025-large-language-diffusion-models/). Preprint, arXiv:2502.09992.
16. Fred Zhangzhi Peng, Zachary Bezemek, Sawan Patel, Jarrid Rector-Brooks, Sherwood Yao, Avishek Joey Bose, Alexander Tong, and Pranam Chatterjee. 2025. Path planning for masked diffusion model sampling. Preprint, arXiv:2502.03540. — [arXiv:2502.03540](https://arxiv.org/abs/2502.03540) [find](https://scholar.google.com/scholar?q=Fred+Zhangzhi+Peng%2C+Zachary+Bezemek%2C+Sawan+Patel%2C+Jarrid+Rector%2DBrooks%2C+Sherwood+Yao%2C+Avishek+Joey+Bose%2C+Alexander+Tong%2C+and+Pranam+Chatterjee%2E+2025%2E+Path+planning+for+masked+diffusion+model+sampling%2E)
17. Naman Jain, King Han, Alex Gu, Wen-Ding Li, Fanjia Yan, Tianjun Zhang, Sida Wang, Armando Solar-Lezama, Koushik Sen, and Ion Stoica. 2024. [Livecodebench: Holistic and contamination free evaluation of large language models for code](https://md.research.elamur.ai/free/en/papers/jain-2024-livecodebench-holistic-contamination-free-evaluation/). Preprint, arXiv:2403.07974.
18. Yuxuan Song, Zheng Zhang, Cheng Luo, Pengyang Gao, Fan Xia, Hao Luo, Zheng Li, Yuehang Yang, Hongli Yu, Xingwei Qu, Yuwei Fu, Jing Su, Ge Zhang, Wenhao Huang, Mingxuan Wang, Lin Yan, Xiaoying Jia, Jingjing Liu, Wei-Ying Ma, and 3 others. 2025. [Seed diffusion: A large-scale diffusion language model w…](https://md.research.elamur.ai/free/en/papers/song-2025-seed-diffusion-large-scale/)
19. Jaeyeon Kim, Lee Cheuk-Kit, Carles Domingo-Enrich, Yilun Du, Sham Kakade, Timothy Ngotiaoco, Sitan Chen, and Michael Albergo. 2025a. Any-order flexible length masked diffusion. Preprint, arXiv:2509.01025. [find](https://scholar.google.com/scholar?q=Jaeyeon+Kim%2C+Lee+Cheuk%2DKit%2C+Carles+Domingo%2DEnrich%2C+Yilun+Du%2C+Sham+Kakade%2C+Timothy+Ngotiaoco%2C+Sitan+Chen%2C+and+Michael+Albergo%2E+2025a%2E+Any%2Dorder+flexible+length+masked+diffusion%2E+Preprint%2C+arXiv%3A2509%2E01)
20. Wen Wang, Bozhen Fang, Chenchen Jing, Yongliang Shen, Yangyi Shen, Qiuyu Wang, Hao Ouyang, Hao Chen, and Chunhua Shen. 2025. Time is a feature: Exploiting temporal dynamics in diffusion language models. Preprint, arXiv:2508.09138. — [arXiv:2508.09138](https://arxiv.org/abs/2508.09138) [find](https://scholar.google.com/scholar?q=Wen+Wang%2C+Bozhen+Fang%2C+Chenchen+Jing%2C+Yongliang+Shen%2C+Yangyi+Shen%2C+Qiuyu+Wang%2C+Hao+Ouyang%2C+Hao+Chen%2C+and+Chunhua+Shen%2E+2025%2E+Time+is+a+feature%3A+Exploiting+temporal+dynamics+in+diffusion+language+model)
21. Jaeyeon Kim, Kulin Shah, Vasilis Kontonis, Sham Kakade, and Sitan Chen. 2025b. Train for the worst, plan for the best: Understanding token ordering in masked diffusions. Preprint, arXiv:2502.06768. — [arXiv:2502.06768](https://arxiv.org/abs/2502.06768) [find](https://scholar.google.com/scholar?q=Jaeyeon+Kim%2C+Kulin+Shah%2C+Vasilis+Kontonis%2C+Sham+Kakade%2C+and+Sitan+Chen%2E+2025b%2E+Train+for+the+worst%2C+plan+for+the+best%3A+Understanding+token+ordering+in+masked+diffusions%2E+Preprint%2C+arXiv%3A2502%2E06768%2E+%E2%80%94+)
22. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. 2023. [Efficient memory management for large language model serving with pagedattention](https://md.research.elamur.ai/free/en/papers/kwon-2023-efficient-memory-management-large-language/). In Proceedings of the 29th Symposium on Operating Systems Principles, pages 611–626.…
23. Xiaoran Liu, Zhigeng Liu, Zengfeng Huang, Qipeng Guo, Ziwei He, and Xipeng Qiu. 2025a. Longllada: Unlocking long context capabilities in diffusion llms. Preprint, arXiv:2506.14429. — [arXiv:2506.14429](https://arxiv.org/abs/2506.14429) [find](https://scholar.google.com/scholar?q=Xiaoran+Liu%2C+Zhigeng+Liu%2C+Zengfeng+Huang%2C+Qipeng+Guo%2C+Ziwei+He%2C+and+Xipeng+Qiu%2E+2025a%2E+Longllada%3A+Unlocking+long+context+capabilities+in+diffusion+llms%2E+Preprint%2C+arXiv%3A2506%2E14429%2E+%E2%80%94+%5BarXiv%3A2506%2E14429)
24. Zhiyuan Liu, Yicun Yang, Yaojie Zhang, Junjie Chen, Chang Zou, Qingyuan Wei, Shaobo Wang, and Linfeng Zhang. 2025b. dllm-cache: Accelerating diffusion large language models with adaptive caching. Preprint, arXiv:2506.06295. — [arXiv:2506.06295](https://arxiv.org/abs/2506.06295) [find](https://scholar.google.com/scholar?q=Zhiyuan+Liu%2C+Yicun+Yang%2C+Yaojie+Zhang%2C+Junjie+Chen%2C+Chang+Zou%2C+Qingyuan+Wei%2C+Shaobo+Wang%2C+and+Linfeng+Zhang%2E+2025b%2E+dllm%2Dcache%3A+Accelerating+diffusion+large+language+models+with+adaptive+caching%2E+Prep)
25. Kaijing Ma, Xinrun Du, Yunran Wang, Haoran Zhang, Zhoufutu Wen, Xingwei Qu, Jian Yang, Jiaheng Liu, Minghao Liu, Xiang Yue, Wenhao Huang, and Ge Zhang. 2025a. Kor-bench: Benchmarking language models on knowledge-orthogonal reasoning tasks. Preprint, arXiv:2410.06526. — [arXiv:2410.06526](https://arxiv.org/abs/2410.06526) [find](https://scholar.google.com/scholar?q=Kaijing+Ma%2C+Xinrun+Du%2C+Yunran+Wang%2C+Haoran+Zhang%2C+Zhoufutu+Wen%2C+Xingwei+Qu%2C+Jian+Yang%2C+Jiaheng+Liu%2C+Minghao+Liu%2C+Xiang+Yue%2C+Wenhao+Huang%2C+and+Ge+Zhang%2E+2025a%2E+Kor%2Dbench%3A+Benchmarking+language+models+o)
26. Qingyan Wei, Yaojie Zhang, Zhiyuan Liu, Dongrui Liu, and Linfeng Zhang. 2025. Accelerating diffusion large language models with slowfast sampling: The three golden principles. Preprint, arXiv:2506.10848. — [arXiv:2506.10848](https://arxiv.org/abs/2506.10848) [find](https://scholar.google.com/scholar?q=Qingyan+Wei%2C+Yaojie+Zhang%2C+Zhiyuan+Liu%2C+Dongrui+Liu%2C+and+Linfeng+Zhang%2E+2025%2E+Accelerating+diffusion+large+language+models+with+slowfast+sampling%3A+The+three+golden+principles%2E+Preprint%2C+arXiv%3A2506%2E108)
27. Chengyue Wu, Hao Zhang, Shuchen Xue, Zhijian Liu, Shizhe Diao, Ligeng Zhu, Ping Luo, Song Han, and Enze Xie. 2025. Fast-dllm: Training-free acceleration of diffusion llm by enabling kv cache and parallel decoding. Preprint, arXiv:2505.22618. — [arXiv:2505.22618](https://arxiv.org/abs/2505.22618) [find](https://scholar.google.com/scholar?q=Chengyue+Wu%2C+Hao+Zhang%2C+Shuchen+Xue%2C+Zhijian+Liu%2C+Shizhe+Diao%2C+Ligeng+Zhu%2C+Ping+Luo%2C+Song+Han%2C+and+Enze+Xie%2E+2025%2E+Fast%2Ddllm%3A+Training%2Dfree+acceleration+of+diffusion+llm+by+enabling+kv+cache+and+paral)
28. Ling Yang, Ye Tian, Bowen Li, Xinchen Zhang, Ke Shen, Yunhai Tong, and Mengdi Wang. 2025. Mmada: Multimodal large diffusion language models. Preprint, arXiv:2505.15809. — [arXiv:2505.15809](https://arxiv.org/abs/2505.15809) [find](https://scholar.google.com/scholar?q=Ling+Yang%2C+Ye+Tian%2C+Bowen+Li%2C+Xinchen+Zhang%2C+Ke+Shen%2C+Yunhai+Tong%2C+and+Mengdi+Wang%2E+2025%2E+Mmada%3A+Multimodal+large+diffusion+language+models%2E+Preprint%2C+arXiv%3A2505%2E15809%2E+%E2%80%94+%5BarXiv%3A2505%2E15809%5D%28https%3A%2F%2Far)
29. Jiacheng Ye, Zhihui Xie, Lin Zheng, Jiahui Gao, Zirui Wu, Xin Jiang, Zhenguo Li, and Lingpeng Kong. 2025. Dream 7b: Diffusion large language models. [find](https://scholar.google.com/scholar?q=Jiacheng+Ye%2C+Zhihui+Xie%2C+Lin+Zheng%2C+Jiahui+Gao%2C+Zirui+Wu%2C+Xin+Jiang%2C+Zhenguo+Li%2C+and+Lingpeng+Kong%2E+2025%2E+Dream+7b%3A+Diffusion+large+language+models%2E)
30. Zebin You, Shen Nie, Xiaolu Zhang, Jun Hu, Jun Zhou, Zhiwu Lu, Ji-Rong Wen, and Chongxuan Li. 2025. Llada-v: [Large language diffusion models](https://md.research.elamur.ai/free/en/papers/nie-2025-large-language-diffusion-models/) with visual instruction tuning. Preprint, arXiv:2505.16933.
31. Runpeng Yu, Xinyin Ma, and Xinchao Wang. 2025. Dimple: Discrete diffusion multimodal large language model with parallel decoding. Preprint, arXiv:2505.16990. — [arXiv:2505.16990](https://arxiv.org/abs/2505.16990) [find](https://scholar.google.com/scholar?q=Runpeng+Yu%2C+Xinyin+Ma%2C+and+Xinchao+Wang%2E+2025%2E+Dimple%3A+Discrete+diffusion+multimodal+large+language+model+with+parallel+decoding%2E+Preprint%2C+arXiv%3A2505%2E16990%2E+%E2%80%94+%5BarXiv%3A2505%2E16990%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs)
32. Fengqi Zhu, Rongzhen Wang, Shen Nie, Xiaolu Zhang, Chunwei Wu, Jun Hu, Jun Zhou, Jianfei Chen, Yankai Lin, Ji-Rong Wen, and Chongxuan Li. 2025a. [Llada 1.5: Variance-reduced preference optimization for large language diffusion models](https://md.research.elamur.ai/free/en/papers/zhu-2026-llada-1-5-variance-reduced/). Preprint, arXiv:2505.19223.
33. Fengqi Zhu, Zebin You, Yipeng Xing, Zenan Huang, Lin Liu, Yihong Zhuang, Guoshan Lu, Kangyu Wang, Xudong Wang, Lanning Wei, Hongrui Guo, Jiaqi Hu, Wentao Ye, Tieyuan Chen, Chenchen Li, Chengfu Tang, Haibo Feng, Jun Hu, Jun Zhou, and 7 others. 2025b. Llada-moe: A sparse moe diffusion language model.… — [arXiv:2509.24389](https://arxiv.org/abs/2509.24389) [find](https://scholar.google.com/scholar?q=Fengqi+Zhu%2C+Zebin+You%2C+Yipeng+Xing%2C+Zenan+Huang%2C+Lin+Liu%2C+Yihong+Zhuang%2C+Guoshan+Lu%2C+Kangyu+Wang%2C+Xudong+Wang%2C+Lanning+Wei%2C+Hongrui+Guo%2C+Jiaqi+Hu%2C+Wentao+Ye%2C+Tieyuan+Chen%2C+Chenchen+Li%2C+Chengfu+Tang%2C+H)

*Updated 1 November 2026*