---
title: "LADR — paper by Wang et al. (2026) — summary and ideas"
description: "The authors propose LADR, a training-free sampling method that requires no retraining. 53 references, cited by 2 papers here."
authors: ["Chenglin Wang", "Yucheng Zhou", "Shuang Chen", "Tao Wang", "Kai Zhang"]
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.1251/"
doi: "10.18653/v1/2026.acl-long.1251"
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-ladr-locality-aware-dynamic-rescue/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/wang-2026-ladr-locality-aware-dynamic-rescue/)

[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/8/) / LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models

# LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models

**Authors:** Chenglin Wang, [Yucheng Zhou](https://md.research.elamur.ai/free/en/authors/yucheng-zhou/), Shuang Chen, [Tao Wang](https://md.research.elamur.ai/free/en/authors/tao-wang/) and [Kai Zhang](https://md.research.elamur.ai/free/en/authors/kai-zhang/) · **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.1251/) · [PDF](https://aclanthology.org/2026.acl-long.1251.pdf) · [DOI 10.18653/v1/2026.acl-long.1251](https://doi.org/10.18653/v1/2026.acl-long.1251) · [OpenAlex](https://openalex.org/W7166797729)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · 2 in our library, of the 149,439 papers whose bibliographies we read

## Summary
Discrete Diffusion Language Models (DLMs) can handle text and images in one framework, but they generate slowly. They typically need 50–100 denoising passes, and bidirectional attention prevents KV-cache reuse. Distillation fixes this only through costly retraining, and speed-up heuristics designed for text ignore the 2D layout of images.

The authors propose LADR, a training-free sampling method that requires no retraining. It relies on spatial locality: a masked VQ token becomes much more predictable once the tokens next to it are known. At each step, LADR does three things:
- It uses morphological operations to find masked tokens bordering already decoded regions, which the authors call the "generation frontier".
- It unmasks ("rescues") only those candidates whose top-1 versus top-2 probability gap is large enough, which limits error propagation.
- It recomputes the diffusion timestep to match the real mask density, using a step the authors call manifold-consistent inverse scheduling.

On four text-to-image benchmarks, LADR is roughly 4× faster than the baseline while keeping, and sometimes improving, quality. Gains are clearest on spatial tasks such as positioning and counting. This makes multimodal DLMs more practical to deploy.

## Key points
- Evaluation covers GenEval, UniGenBench, DPG-Bench and T2I-CompBench. Comparisons include the Lumina-DiMOO backbone and the ML-Cache and Prophet acceleration baselines.
- Inference time drops from about 57 s to about 13 s per image. This is roughly 4× faster than non-cached sampling and about 2× faster than heuristic methods.
- In the GenEval ablation, ranking candidates by the confidence margin scores 87.30 overall. Plain top-1 confidence scores 85.24 and random neighbor selection scores 83.74.
- Compared with top-1 confidence, the margin criterion helps most on structural categories: Position rises from 80.25 to 85.50 and Attribute from 71.25 to 74.75.
- Stated limitations:
  - The hyperparameters (confidence threshold τ and rescue ratio α) were set empirically rather than optimized.
  - Output quality depends on the underlying backbone model.

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

## Institutions

- [Zhejiang University](https://md.research.elamur.ai/free/en/universities/zhejiang-university/) — University, Hangzhou, China · 172 works · 12,988 citations
- [Nanjing University](https://md.research.elamur.ai/free/en/universities/nanjing-university/) — University, Nanjing, China · 74 works · 7,814 citations
- [East China Normal University](https://md.research.elamur.ai/free/en/universities/east-china-normal-university/) — University, Shanghai, China · 50 works · 2,174 citations
- [University of Macau](https://md.research.elamur.ai/free/en/universities/university-of-macau/) — University, Macao, China · 33 works · 2,223 citations

📚 Shelves (2): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 2,869 works

🧭 Topics (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works · [🖼️ Vision & language](https://md.research.elamur.ai/free/en/topics/multimodal/) · 657 works

## Cite

Cite as: Chenglin Wang, Yucheng Zhou, Shuang Chen, Tao Wang and Kai Zhang. 2026. LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models. 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{wang2026ladr,
  title = {{LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models}},
  author = {Chenglin Wang, Yucheng Zhou, Shuang Chen, Tao Wang and Kai Zhang},
  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.1251},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1251}
}
```

- [← previous Label and Explanation Variation in LLM-Based Annotation: a Case Study in Natural Language Inference](https://md.research.elamur.ai/free/en/papers/kulmizev-2026-label-explanation-variation-llm-based/)
- [next → Language of Thought Shapes Output Diversity in Large Language Models](https://md.research.elamur.ai/free/en/papers/xu-2026-language-thought-shapes-output-diversity/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/8/)

## References

53 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, and 1 others. 2023. [Gpt-4 technical report](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/). arXiv preprint arXiv:2303.08774.
2. Marianne Arriola, Subham Sekhar Sahoo, Aaron Gokaslan, Zhihan Yang, Zhixuan Qi, Jiaqi Han, Justin T Chiu, and Volodymyr Kuleshov. 2025. Block diffusion: Interpolating between autoregressive and diffusion language models. In Proceedings of International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Marianne+Arriola%2C+Subham+Sekhar+Sahoo%2C+Aaron+Gokaslan%2C+Zhihan+Yang%2C+Zhixuan+Qi%2C+Jiaqi+Han%2C+Justin+T+Chiu%2C+and+Volodymyr+Kuleshov%2E+2025%2E+Block+diffusion%3A+Interpolating+between+autoregressive+and+diffus)
3. Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, and 1 others. 2023. [Qwen technical report](https://md.research.elamur.ai/free/en/papers/bai-2023-qwen-technical-report/). arXiv preprint arXiv:2309.16609.
4. Zefan Cai, Yichi Zhang, Bofei Gao, Yuliang Liu, Tianyu Liu, Keming Lu, Wayne Xiong, Yue Dong, Baobao Chang, Junjie Hu, and Xiao Wen. 2024. [Pyramidkv: Dynamic kv cache compression based on pyramidal information funneling](https://md.research.elamur.ai/free/en/papers/cai-2024-pyramidkv-dynamic-kv-cache-compression/). arXiv preprint arXiv:2406.02069.
5. Huiwen Chang, Han Zhang, Jarred Barber, AJ Maschinot, José Lezama, Lu Jiang, Ming-Hsuan Yang, Kevin Murphy, William T Freeman, Michael Rubinstein, and 1 others. 2023. Muse: Text-to-image generation via masked generative transformers. In Proceedings of International Conference on Machine Learning, pa… [find](https://scholar.google.com/scholar?q=Huiwen+Chang%2C+Han+Zhang%2C+Jarred+Barber%2C+AJ+Maschinot%2C+Jos%C3%A9+Lezama%2C+Lu+Jiang%2C+Ming%2DHsuan+Yang%2C+Kevin+Murphy%2C+William+T+Freeman%2C+Michael+Rubinstein%2C+and+1+others%2E+2023%2E+Muse%3A+Text%2Dto%2Dimage+generation+vi)
6. Huiwen Chang, Han Zhang, Lu Jiang, Ce Liu, and William T Freeman. 2022. Maskgit: Masked generative image transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11315–11325. [find](https://scholar.google.com/scholar?q=Huiwen+Chang%2C+Han+Zhang%2C+Lu+Jiang%2C+Ce+Liu%2C+and+William+T+Freeman%2E+2022%2E+Maskgit%3A+Masked+generative+image+transformer%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Computer+Vision+and+Pattern+Recogniti)
7. Guanjie Chen, Xinyu Zhao, Yucheng Zhou, Xiaoye Qu, Tianlong Chen, and Yu Cheng. 2025. Towards stabilized and efficient diffusion transformers through long-skip-connections with spectral constraints. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 17708–17718. [find](https://scholar.google.com/scholar?q=Guanjie+Chen%2C+Xinyu+Zhao%2C+Yucheng+Zhou%2C+Xiaoye+Qu%2C+Tianlong+Chen%2C+and+Yu+Cheng%2E+2025%2E+Towards+stabilized+and+efficient+diffusion+transformers+through+long%2Dskip%2Dconnections+with+spectral+constraints%2E+I)
8. Justin Deschenaux and Caglar Gulcehre. 2025. Beyond autoregression: Fast LLMs via self-distillation through time. In Proceedings of International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Justin+Deschenaux+and+Caglar+Gulcehre%2E+2025%2E+Beyond+autoregression%3A+Fast+LLMs+via+self%2Ddistillation+through+time%2E+In+Proceedings+of+International+Conference+on+Learning+Representations%2E)
9. Patrick Esser, Robin Rombach, and Bjorn Ommer. 2021. Taming transformers for high-resolution image synthesis. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12873–12883. [find](https://scholar.google.com/scholar?q=Patrick+Esser%2C+Robin+Rombach%2C+and+Bjorn+Ommer%2E+2021%2E+Taming+transformers+for+high%2Dresolution+image+synthesis%2E+In+Proceedings+of+the+IEEE%2FCVF+conference+on+computer+vision+and+pattern+recognition%2C+page)
10. Dhruba Ghosh, Hannaneh Hajishirzi, and Ludwig Schmidt. 2023. Geneval: An object-focused framework for evaluating text-to-image alignment. In Proceedings of Advances in Neural Information Processing Systems, pages 52132–52152. [find](https://scholar.google.com/scholar?q=Dhruba+Ghosh%2C+Hannaneh+Hajishirzi%2C+and+Ludwig+Schmidt%2E+2023%2E+Geneval%3A+An+object%2Dfocused+framework+for+evaluating+text%2Dto%2Dimage+alignment%2E+In+Proceedings+of+Advances+in+Neural+Information+Processing+Sy)
11. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, and 1 others. 2025. [Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-r1-incentivizing-reasoning-capability/). arXiv preprint arXiv:2501.12948.
12. Satoshi Hayakawa, Yuhta Takida, Masaaki Imaizumi, Hiromi Wakaki, and Yuki Mitsufuji. 2025. Distillation of discrete diffusion through dimensional correlations. In Proceedings of International Conference on Machine Learning. [find](https://scholar.google.com/scholar?q=Satoshi+Hayakawa%2C+Yuhta+Takida%2C+Masaaki+Imaizumi%2C+Hiromi+Wakaki%2C+and+Yuki+Mitsufuji%2E+2025%2E+Distillation+of+discrete+diffusion+through+dimensional+correlations%2E+In+Proceedings+of+International+Conferen)
13. G Hinton. 2014. [Distilling the knowledge in a neural network](https://md.research.elamur.ai/free/en/papers/hinton-2015-distilling-knowledge-neural-network/). In Deep Learning and Representation Learning Workshop in Conjunction with NIPS. 27178 9
14. Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. In Proceedings of Advances in Neural Information Processing Systems, pages 6840–6851. [find](https://scholar.google.com/scholar?q=Jonathan+Ho%2C+Ajay+Jain%2C+and+Pieter+Abbeel%2E+2020%2E+Denoising+diffusion+probabilistic+models%2E+In+Proceedings+of+Advances+in+Neural+Information+Processing+Systems%2C+pages+6840%E2%80%936851%2E)
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/). arXiv preprint arXiv:2502.09992.
16. Xiwei Hu, Rui Wang, Yixiao Fang, Bin Fu, Pei Cheng, and Gang Yu. 2024. Ella: Equip diffusion models with llm for enhanced semantic alignment. arXiv preprint arXiv:2403.05135. — [arXiv:2403.05135](https://arxiv.org/abs/2403.05135) [find](https://scholar.google.com/scholar?q=Xiwei+Hu%2C+Rui+Wang%2C+Yixiao+Fang%2C+Bin+Fu%2C+Pei+Cheng%2C+and+Gang+Yu%2E+2024%2E+Ella%3A+Equip+diffusion+models+with+llm+for+enhanced+semantic+alignment%2E+arXiv+preprint+arXiv%3A2403%2E05135%2E+%E2%80%94+%5BarXiv%3A2403%2E05135%5D%28http)
17. Alec Radford, Karthik Narasimhan, Tim Salimans, Ilya Sutskever, and 1 others. 2018. Improving language understanding by generative pre-training. [find](https://scholar.google.com/scholar?q=Alec+Radford%2C+Karthik+Narasimhan%2C+Tim+Salimans%2C+Ilya+Sutskever%2C+and+1+others%2E+2018%2E+Improving+language+understanding+by+generative+pre%2Dtraining%2E)
18. Zhanqiu Hu, Jian Meng, Yash Akhauri, Mohamed S Abdelfattah, Jae-sun Seo, Zhiru Zhang, and Udit Gupta. 2025. Accelerating diffusion language model inference via efficient kv caching and guided diffusion. arXiv preprint arXiv:2505.21467. — [arXiv:2505.21467](https://arxiv.org/abs/2505.21467) [find](https://scholar.google.com/scholar?q=Zhanqiu+Hu%2C+Jian+Meng%2C+Yash+Akhauri%2C+Mohamed+S+Abdelfattah%2C+Jae%2Dsun+Seo%2C+Zhiru+Zhang%2C+and+Udit+Gupta%2E+2025%2E+Accelerating+diffusion+language+model+inference+via+efficient+kv+caching+and+guided+diffusio)
19. Kaiyi Huang, Kaiyue Sun, Enze Xie, Zhenguo Li, and Xihui Liu. 2023. T2i-compbench: A comprehensive benchmark for open-world compositional text-to-image generation. In Proceedings of Advances in Neural Information Processing Systems, pages 78723–78747. [find](https://scholar.google.com/scholar?q=Kaiyi+Huang%2C+Kaiyue+Sun%2C+Enze+Xie%2C+Zhenguo+Li%2C+and+Xihui+Liu%2E+2023%2E+T2i%2Dcompbench%3A+A+comprehensive+benchmark+for+open%2Dworld+compositional+text%2Dto%2Dimage+generation%2E+In+Proceedings+of+Advances+in+Neural)
20. Daniel Mingyi Israel, Guy Van den Broeck, and Aditya Grover. 2025. Accelerating diffusion LLMs via adaptive parallel decoding. In Proceedings of Annual Conference on Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Daniel+Mingyi+Israel%2C+Guy+Van+den+Broeck%2C+and+Aditya+Grover%2E+2025%2E+Accelerating+diffusion+LLMs+via+adaptive+parallel+decoding%2E+In+Proceedings+of+Annual+Conference+on+Neural+Information+Processing+Syst)
21. Dongjun Kim, Chieh-Hsin Lai, Wei-Hsiang Liao, Naoki Murata, Yuhta Takida, Toshimitsu Uesaka, Yutong He, Yuki Mitsufuji, and Stefano Ermon. 2024. Consistency trajectory models: Learning probability flow ODE trajectory of diffusion. In Proceedings of International Conference on Learning Representation… [find](https://scholar.google.com/scholar?q=Dongjun+Kim%2C+Chieh%2DHsin+Lai%2C+Wei%2DHsiang+Liao%2C+Naoki+Murata%2C+Yuhta+Takida%2C+Toshimitsu+Uesaka%2C+Yutong+He%2C+Yuki+Mitsufuji%2C+and+Stefano+Ermon%2E+2024%2E+Consistency+trajectory+models%3A+Learning+probability+flo)
22. Pengxiang Li, Yefan Zhou, Dilxat Muhtar, Lu Yin, Shilin Yan, Li Shen, Yi Liang, Soroush Vosoughi, and Shiwei Liu. 2025a. Diffusion language models know the answer before decoding. arXiv preprint arXiv:2508.19982. — [arXiv:2508.19982](https://arxiv.org/abs/2508.19982) [find](https://scholar.google.com/scholar?q=Pengxiang+Li%2C+Yefan+Zhou%2C+Dilxat+Muhtar%2C+Lu+Yin%2C+Shilin+Yan%2C+Li+Shen%2C+Yi+Liang%2C+Soroush+Vosoughi%2C+and+Shiwei+Liu%2E+2025a%2E+Diffusion+language+models+know+the+answer+before+decoding%2E+arXiv+preprint+arXiv)
23. Shufan Li, Jiuxiang Gu, Kangning Liu, Zhe Lin, Zijun Wei, Aditya Grover, and Jason Kuen. 2025b. Lavida-o: Elastic large masked diffusion models for unified multimodal understanding and generation. arXiv preprint arXiv:2509.19244. — [arXiv:2509.19244](https://arxiv.org/abs/2509.19244) [find](https://scholar.google.com/scholar?q=Shufan+Li%2C+Jiuxiang+Gu%2C+Kangning+Liu%2C+Zhe+Lin%2C+Zijun+Wei%2C+Aditya+Grover%2C+and+Jason+Kuen%2E+2025b%2E+Lavida%2Do%3A+Elastic+large+masked+diffusion+models+for+unified+multimodal+understanding+and+generation%2E+arX)
24. Shufan Li, Konstantinos Kallidromitis, Hritik Bansal, Akash Gokul, Yusuke Kato, Kazuki Kozuka, Jason Kuen, Zhe Lin, Kai-Wei Chang, and Aditya Grover. 2025c. Lavida: A large diffusion language model for multimodal understanding. arXiv preprint arXiv:2505.16839. — [arXiv:2505.16839](https://arxiv.org/abs/2505.16839) [find](https://scholar.google.com/scholar?q=Shufan+Li%2C+Konstantinos+Kallidromitis%2C+Hritik+Bansal%2C+Akash+Gokul%2C+Yusuke+Kato%2C+Kazuki+Kozuka%2C+Jason+Kuen%2C+Zhe+Lin%2C+Kai%2DWei+Chang%2C+and+Aditya+Grover%2E+2025c%2E+Lavida%3A+A+large+diffusion+language+model+fo)
25. Yuhong Li, Yingbing Huang, Bowen Yang, Bharat Venkitesh, Acyr Locatelli, Hanchen Ye, Tianle Cai, Patrick Lewis, and Deming Chen. 2024. [Snapkv: Llm knows what you are looking for before generation](https://md.research.elamur.ai/free/en/papers/li-2024-snapkv-llm-knows-looking-before/). In Proceedings of Advances in Neural Information Processing Systems, pages 22947–22970.
26. Xingchao Liu, Chengyue Gong, and 1 others. 2023. [Flow straight and fast: Learning to generate and transfer data with rectified flow](https://md.research.elamur.ai/free/en/papers/liu-2022-flow-straight-fast-learning-generate/). In The Eleventh International Conference on Learning Representations.
27. Dominic Rampas, Pablo Pernias, and Marc Aubre-ville. 2022. A novel sampling scheme for text-and image-conditional image synthesis in quantized latent spaces. arXiv preprint arXiv:2211.07292. — [arXiv:2211.07292](https://arxiv.org/abs/2211.07292) [find](https://scholar.google.com/scholar?q=Dominic+Rampas%2C+Pablo+Pernias%2C+and+Marc+Aubre%2Dville%2E+2022%2E+A+novel+sampling+scheme+for+text%2Dand+image%2Dconditional+image+synthesis+in+quantized+latent+spaces%2E+arXiv+preprint+arXiv%3A2211%2E07292%2E+%E2%80%94+%5BarXiv%3A)
28. Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10684–10695. [find](https://scholar.google.com/scholar?q=Robin+Rombach%2C+Andreas+Blattmann%2C+Dominik+Lorenz%2C+Patrick+Esser%2C+and+Bj%C3%B6rn+Ommer%2E+2022%2E+High%2Dresolution+image+synthesis+with+latent+diffusion+models%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Compu)
29. Subham Sahoo, Marianne Arriola, Yair Schiff, Aaron Gokaslan, Edgar Marroquin, Justin Chiu, Alexander Rush, and Volodymyr Kuleshov. 2024. [Simple and effective masked diffusion language models](https://md.research.elamur.ai/free/en/papers/sahoo-2024-simple-effective-masked-diffusion-language/). In Proceedings of Advances in Neural Information Processing Systems, pages 130136–130184.
30. Han Song, Yucheng Zhou, Jianbing Shen, and Yu Cheng. 2026. From broad exploration to stable synthesis: Entropy-guided optimization for autoregressive image generation. In The Fourteenth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Han+Song%2C+Yucheng+Zhou%2C+Jianbing+Shen%2C+and+Yu+Cheng%2E+2026%2E+From+broad+exploration+to+stable+synthesis%3A+Entropy%2Dguided+optimization+for+autoregressive+image+generation%2E+In+The+Fourteenth+International+)
31. Yang Song, Prafulla Dhariwal, Mark Chen, and Ilya Sutskever. 2023. Consistency models. In Proceedings of International Conference on Machine Learning, pages 32211–32252. [find](https://scholar.google.com/scholar?q=Yang+Song%2C+Prafulla+Dhariwal%2C+Mark+Chen%2C+and+Ilya+Sutskever%2E+2023%2E+Consistency+models%2E+In+Proceedings+of+International+Conference+on+Machine+Learning%2C+pages+32211%E2%80%9332252%2E)
32. Yuxuan Song, Zheng Zhang, Cheng Luo, Pengyang Gao, Fan Xia, Hao Luo, Zheng Li, Yuehang Yang, Hongli Yu, Xingwei Qu, and 1 others. 2025. [Seed diffusion: A large-scale diffusion language model with high-speed inference](https://md.research.elamur.ai/free/en/papers/song-2025-seed-diffusion-large-scale/). arXiv preprint arXiv:2508.02193.
33. Alexander Swerdlow, Mihir Prabhudesai, Siddharth Gandhi, Deepak Pathak, and Katerina Fragkiadaki. 2025. Unified multimodal discrete diffusion. arXiv preprint arXiv:2503.20853. — [arXiv:2503.20853](https://arxiv.org/abs/2503.20853) [find](https://scholar.google.com/scholar?q=Alexander+Swerdlow%2C+Mihir+Prabhudesai%2C+Siddharth+Gandhi%2C+Deepak+Pathak%2C+and+Katerina+Fragkiadaki%2E+2025%2E+Unified+multimodal+discrete+diffusion%2E+arXiv+preprint+arXiv%3A2503%2E20853%2E+%E2%80%94+%5BarXiv%3A2503%2E20853%5D%28htt)
34. Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, and 1 others. 2023. [Llama: Open and efficient foundation language models](https://md.research.elamur.ai/free/en/papers/touvron-2023-llama-open-efficient-foundation-language/). arXiv preprint arXiv:2302.13971.
35. Chenglin Wang, Yucheng Zhou, Qianning Wang, Zhe Wang, and Kai Zhang. 2025a. Complexbench-edit: Benchmarking complex instruction-driven image editing via compositional dependencies. In Proceedings of the 33rd ACM International Conference on Multimedia, pages 13391–13397. 27179 10 [find](https://scholar.google.com/scholar?q=Chenglin+Wang%2C+Yucheng+Zhou%2C+Qianning+Wang%2C+Zhe+Wang%2C+and+Kai+Zhang%2E+2025a%2E+Complexbench%2Dedit%3A+Benchmarking+complex+instruction%2Ddriven+image+editing+via+compositional+dependencies%2E+In+Proceedings+of+t)
36. Chenglin Wang, Yucheng Zhou, Zhe Wang, Zijie Zhai, Jianbing Shen, and Kai Zhang. 2025b. Alternate geometric and semantic denoising diffusion for protein inverse folding. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pages 350–366. Springer. [find](https://scholar.google.com/scholar?q=Chenglin+Wang%2C+Yucheng+Zhou%2C+Zhe+Wang%2C+Zijie+Zhai%2C+Jianbing+Shen%2C+and+Kai+Zhang%2E+2025b%2E+Alternate+geometric+and+semantic+denoising+diffusion+for+protein+inverse+folding%2E+In+Joint+European+Conference+o)
37. Xu Wang, Chenkai Xu, Yijie Jin, Jiachun Jin, Hao Zhang, and Zhijie Deng. 2025c. Diffusion llms can do faster-than-ar inference via discrete diffusion forcing. arXiv preprint arXiv:2508.09192. — [arXiv:2508.09192](https://arxiv.org/abs/2508.09192) [find](https://scholar.google.com/scholar?q=Xu+Wang%2C+Chenkai+Xu%2C+Yijie+Jin%2C+Jiachun+Jin%2C+Hao+Zhang%2C+and+Zhijie+Deng%2E+2025c%2E+Diffusion+llms+can+do+faster%2Dthan%2Dar+inference+via+discrete+diffusion+forcing%2E+arXiv+preprint+arXiv%3A2508%2E09192%2E+%E2%80%94+%5BarXiv)
38. Yibin Wang, Zhimin Li, Yuhang Zang, Yujie Zhou, Jiazi Bu, Chunyu Wang, Qinglin Lu, Cheng Jin, and Jiaqi Wang. 2025d. Pref-grpo: Pairwise preference reward-based grpo for stable text-to-image reinforcement learning. arXiv preprint arXiv:2508.20751. — [arXiv:2508.20751](https://arxiv.org/abs/2508.20751) [find](https://scholar.google.com/scholar?q=Yibin+Wang%2C+Zhimin+Li%2C+Yuhang+Zang%2C+Yujie+Zhou%2C+Jiazi+Bu%2C+Chunyu+Wang%2C+Qinglin+Lu%2C+Cheng+Jin%2C+and+Jiaqi+Wang%2E+2025d%2E+Pref%2Dgrpo%3A+Pairwise+preference+reward%2Dbased+grpo+for+stable+text%2Dto%2Dimage+reinforce)
39. Chengyue Wu, Hao Zhang, Shuchen Xue, Shizhe Diao, Yonggan Fu, Zhijian Liu, Pavlo Molchanov, Ping Luo, Song Han, and Enze Xie. 2025a. Fast-dllm v2: Efficient block-diffusion llm. arXiv preprint arXiv:2509.26328. — [arXiv:2509.26328](https://arxiv.org/abs/2509.26328) [find](https://scholar.google.com/scholar?q=Chengyue+Wu%2C+Hao+Zhang%2C+Shuchen+Xue%2C+Shizhe+Diao%2C+Yonggan+Fu%2C+Zhijian+Liu%2C+Pavlo+Molchanov%2C+Ping+Luo%2C+Song+Han%2C+and+Enze+Xie%2E+2025a%2E+Fast%2Ddllm+v2%3A+Efficient+block%2Ddiffusion+llm%2E+arXiv+preprint+arXiv%3A2)
40. Chengyue Wu, Hao Zhang, Shuchen Xue, Zhijian Liu, Shizhe Diao, Ligeng Zhu, Ping Luo, Song Han, and Enze Xie. 2025b. Fast-dllm: Training-free acceleration of diffusion llm by enabling kv cache and parallel decoding. arXiv 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+2025b%2E+Fast%2Ddllm%3A+Training%2Dfree+acceleration+of+diffusion+llm+by+enabling+kv+cache+and+para)
41. Yi Xin, Qi Qin, Siqi Luo, Kaiwen Zhu, Juncheng Yan, Yan Tai, Jiayi Lei, Yuewen Cao, Keqi Wang, Yibin Wang, and 1 others. 2025. Lumina-dimoo: An omni diffusion large language model for multimodal generation and understanding. arXiv preprint arXiv:2510.06308. — [arXiv:2510.06308](https://arxiv.org/abs/2510.06308) [find](https://scholar.google.com/scholar?q=Yi+Xin%2C+Qi+Qin%2C+Siqi+Luo%2C+Kaiwen+Zhu%2C+Juncheng+Yan%2C+Yan+Tai%2C+Jiayi+Lei%2C+Yuewen+Cao%2C+Keqi+Wang%2C+Yibin+Wang%2C+and+1+others%2E+2025%2E+Lumina%2Ddimoo%3A+An+omni+diffusion+large+language+model+for+multimodal+gener)
42. Hongji Yang, Wencheng Han, Yucheng Zhou, and Jianbing Shen. 2025a. Dc-controlnet: Decoupling inter-and intra-element conditions in image generation with diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 19065–19074. the IEEE/CVF Conference on Compute… [find](https://scholar.google.com/scholar?q=Hongji+Yang%2C+Wencheng+Han%2C+Yucheng+Zhou%2C+and+Jianbing+Shen%2E+2025a%2E+Dc%2Dcontrolnet%3A+Decoupling+inter%2Dand+intra%2Delement+conditions+in+image+generation+with+diffusion+models%2E+In+Proceedings+of+the+IEEE%2FCV)
43. 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. arXiv preprint arXiv:2505.16933.
44. Yucheng Zhou, Hao Li, and Jianbing Shen. 2026. Condition errors refinement in autoregressive image generation with diffusion loss. In The Fourteenth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Yucheng+Zhou%2C+Hao+Li%2C+and+Jianbing+Shen%2E+2026%2E+Condition+errors+refinement+in+autoregressive+image+generation+with+diffusion+loss%2E+In+The+Fourteenth+International+Conference+on+Learning+Representation)
45. Yucheng Zhou, Jiahao Yuan, and Qianning Wang. 2025a. Draw all your imagine: A holistic benchmark and agent framework for complex instruction-based image generation. arXiv preprint arXiv:2505.24787. — [arXiv:2505.24787](https://arxiv.org/abs/2505.24787) [find](https://scholar.google.com/scholar?q=Yucheng+Zhou%2C+Jiahao+Yuan%2C+and+Qianning+Wang%2E+2025a%2E+Draw+all+your+imagine%3A+A+holistic+benchmark+and+agent+framework+for+complex+instruction%2Dbased+image+generation%2E+arXiv+preprint+arXiv%3A2505%2E24787%2E+%E2%80%94+)
46. Yucheng Zhou, Jihai Zhang, Guanjie Chen, Jianbing Shen, and Yu Cheng. 2024. Less is more: Vision representation compression for efficient video generation with large language models. OpenReview. [find](https://scholar.google.com/scholar?q=Yucheng+Zhou%2C+Jihai+Zhang%2C+Guanjie+Chen%2C+Jianbing+Shen%2C+and+Yu+Cheng%2E+2024%2E+Less+is+more%3A+Vision+representation+compression+for+efficient+video+generation+with+large+language+models%2E+OpenReview%2E)
47. Yucheng Zhou, Huan Zheng, Dubing Chen, Hongji Yang, Wencheng Han, and Jianbing Shen. 2025b. From medical llms to versatile medical agents: A comprehensive survey. [find](https://scholar.google.com/scholar?q=Yucheng+Zhou%2C+Huan+Zheng%2C+Dubing+Chen%2C+Hongji+Yang%2C+Wencheng+Han%2C+and+Jianbing+Shen%2E+2025b%2E+From+medical+llms+to+versatile+medical+agents%3A+A+comprehensive+survey%2E)
48. Yuanzhi Zhu, Xi Wang, Stéphane Lathuilière, and Vicky Kalogeiton. 2025a. Di [m] o: Distilling masked diffusion models into one-step generator. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 18606–18618. [find](https://scholar.google.com/scholar?q=Yuanzhi+Zhu%2C+Xi+Wang%2C+St%C3%A9phane+Lathuili%C3%A8re%2C+and+Vicky+Kalogeiton%2E+2025a%2E+Di+%5Bm%5D+o%3A+Distilling+masked+diffusion+models+into+one%2Dstep+generator%2E+In+Proceedings+of+the+IEEE%2FCVF+International+Conference+o)
49. Yuanzhi Zhu, Xi Wang, Stéphane Lathuilière, and Vicky Kalogeiton. 2025b. Soft-di [m] o: Improving one-step discrete image generation with soft embeddings. arXiv preprint arXiv:2509.22925. — [arXiv:2509.22925](https://arxiv.org/abs/2509.22925) [find](https://scholar.google.com/scholar?q=Yuanzhi+Zhu%2C+Xi+Wang%2C+St%C3%A9phane+Lathuili%C3%A8re%2C+and+Vicky+Kalogeiton%2E+2025b%2E+Soft%2Ddi+%5Bm%5D+o%3A+Improving+one%2Dstep+discrete+image+generation+with+soft+embeddings%2E+arXiv+preprint+arXiv%3A2509%2E22925%2E+%E2%80%94+%5BarXiv%3A250)
50. Hongji Yang, Yucheng Zhou, Wencheng Han, and Jianbing Shen. 2025b. Self-rewarding large vision-language models for optimizing prompts in text-to-image generation. In Findings of the Association for Computational Linguistics: ACL 2025, pages 7332– 7349. [find](https://scholar.google.com/scholar?q=Hongji+Yang%2C+Yucheng+Zhou%2C+Wencheng+Han%2C+and+Jianbing+Shen%2E+2025b%2E+Self%2Drewarding+large+vision%2Dlanguage+models+for+optimizing+prompts+in+text%2Dto%2Dimage+generation%2E+In+Findings+of+the+Association+for+Co)
51. Hongji Yang, Yucheng Zhou, Wencheng Han, Runzhou Tao, Zhongying Qiu, Jianfei Yang, and Jianbing Shen. 2025c. Hicogen: Hierarchical compositional text-to-image generation in diffusion models via reinforcement learning. arXiv preprint arXiv:2511.19965. — [arXiv:2511.19965](https://arxiv.org/abs/2511.19965) [find](https://scholar.google.com/scholar?q=Hongji+Yang%2C+Yucheng+Zhou%2C+Wencheng+Han%2C+Runzhou+Tao%2C+Zhongying+Qiu%2C+Jianfei+Yang%2C+and+Jianbing+Shen%2E+2025c%2E+Hicogen%3A+Hierarchical+compositional+text%2Dto%2Dimage+generation+in+diffusion+models+via+reinfo)
52. Jiacheng Ye, Zhihui Xie, Lin Zheng, Jiahui Gao, Zirui Wu, Xin Jiang, Zhenguo Li, and Lingpeng Kong. 2025. Dream 7b: Diffusion large language models. arXiv preprint arXiv:2508.15487. [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+arXiv+preprint+arXiv%3A2508%2E15487%2E)
53. Tianwei Yin, Michaël Gharbi, Richard Zhang, Eli Shechtman, Fredo Durand, William T Freeman, and Taesung Park. 2024. One-step diffusion with distribution matching distillation. In Proceedings of 27180 11 [find](https://scholar.google.com/scholar?q=Tianwei+Yin%2C+Micha%C3%ABl+Gharbi%2C+Richard+Zhang%2C+Eli+Shechtman%2C+Fredo+Durand%2C+William+T+Freeman%2C+and+Taesung+Park%2E+2024%2E+One%2Dstep+diffusion+with+distribution+matching+distillation%2E+In+Proceedings+of+27180+)

*Updated 31 October 2026*