Elamur Research
☰
home / papers / 2025 / Large Language Diffusion Models

Large Language Diffusion Models

Written for the Research library (claude-opus-5-5) from the paper text, in our own words: the license "arXiv — non-exclusive license to distribute" does not allow redistributing the text itself. Original — arxiv.org/pdf/2502.09992

Summary

Modern LLMs almost all rely on autoregressive next-token prediction. Many assume that scaling, in-context learning and instruction-following depend on this design. The authors test that assumption with LLaDA, a masked diffusion language model trained from scratch.

During training, each token in a sequence is hidden with a probability drawn uniformly from 0 to 1. A Transformer without a causal mask then learns to reconstruct the hidden tokens. The resulting loss upper-bounds the negative log-likelihood, so LLaDA remains a principled generative model. To generate, it starts from a fully masked answer, predicts all masked positions in parallel and refines them over several steps with remasking.

LLaDA 8B was pre-trained on 2.3T tokens and then given SFT. It:

  • matches self-built autoregressive baselines trained on the same data;
  • rivals LLaMA3 8B on few-shot benchmarks;
  • holds multi-turn dialogues;
  • avoids the reversal curse.

The study suggests that core LLM abilities come from generative modeling principles, not from left-to-right factorization. This makes diffusion a credible alternative paradigm.

Key points

  • Training scale: LLaDA 8B (plus a 1B variant) was pre-trained on 2.3T tokens with 4096-token sequences, costing about 0.13M H800 GPU hours. SFT used 4.5M prompt–response pairs, masking only response tokens.
  • Architecture vs. LLaMA3 8B: LLaDA uses plain multi-head attention instead of grouped-query attention, since KV caching does not apply. It compensates with a smaller FFN and uses its own tokenizer. The masking ratio is random rather than fixed as in BERT.
  • Scaling and benchmarks: Up to 10^23 FLOPs, LLaDA tracks ARM baselines on six tasks, including MMLU and GSM8K. LLaDA 8B Base beats LLaMA2 7B Base on nearly all of 15 zero/few-shot tasks and is roughly on par with LLaMA3 8B Base.
  • Reversal curse: Performance is consistent on forward and reversed tasks, and LLaDA outperforms GPT-4o on reversal poem completion.
  • Limitations: Generation length must be set by the user. There is no RL alignment and no KV cache or other system-level optimizations. Like-for-like ARM comparisons were capped below 10^23 FLOPs.

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

Fields: Other

Cited by

141 papers of our library cite it (of the 140673 whose bibliographies we read).

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2502.09992 · arXiv 2502.09992

@misc{nie2025large,
  title = {{Large Language Diffusion Models}},
  author = {Nie, Shen and Zhu, Fengqi and You, Zebin and Zhang, Xiaolu and Ou, Jingyang and Hu, Jun and Zhou, Jun and Lin, Yankai and Wen, Ji-Rong and Li, Chongxuan},
  year = {2025},
  doi = {10.48550/arxiv.2502.09992},
  eprint = {2502.09992},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2502.09992}
}

References

125 references, 26 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, et al. A survey of large language models. arXiv preprint arXiv:2303.18223, 2023. find our card
  2. Alec Radford. Improving language understanding by generative pre-training, 2018. find
  3. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al. Language models are unsupervised multitask learners. OpenAI blog, 1(8):9, 2019. find
  4. Tom B Brown. Language models are few-shot learners. arXiv preprint arXiv:2005.14165, 2020. find
  5. OpenAI. ChatGPT: Optimizing Language Models for Dialogue. OpenAI blog, November 2022. URL https://openai.com/blog/chatgpt/. ↗
  6. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783, 2024. find
  7. Ashish Vaswani. Attention is all you need. arXiv preprint arXiv:1706.03762, 2017. find
  8. Ronald A Fisher. On the mathematical foundations of theoretical statistics. Philosophical transactions of the Royal Society of London. Series A, containing papers of a mathematical or physical character, 222(594-604):309–368, 1922. find
  9. Fan Bao, Shen Nie, Kaiwen Xue, Yue Cao, Chongxuan Li, Hang Su, and Jun Zhu. All are worth words: A vit backbone for diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 22669–22679, 2023. find
  10. William Peebles and Saining Xie. Scalable diffusion models with transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 4195–4205, 2023. find our card
  11. Tim Brooks, Bill Peebles, Connor Holmes, Will DePue, Yufei Guo, Li Jing, David Schnurr, Joe Taylor, Troy Luhman, Eric Luhman, Clarence Ng, Ricky Wang, and Aditya Ramesh. Video generation models as world simulators. 2024. URL https://openai.com/research/video-generation-models-as-world-simulators. ↗
  12. Gregoire Deletang, Anian Ruoss, Paul-Ambroise Duquenne, Elliot Catt, Tim Genewein, Christopher Mattern, Jordi Grau-Moya, Li Kevin Wenliang, Matthew Aitchison, Laurent Orseau, et al. Language modeling is compression. In The Twelfth International Conference on Learning Representations. find
  13. Yuzhen Huang, Jinghan Zhang, Zifei Shan, and Junxian He. Compression represents intelligence linearly. arXiv preprint arXiv:2404.09937, 2024. — [arXiv:2404.09937](https://arxiv.org/abs/2404.09937) find
  14. Claude Elwood Shannon. A mathematical theory of communication. The Bell system technical journal, 27(3):379–423, 1948. find
  15. Lukas Berglund, Meg Tong, Max Kaufmann, Mikita Balesni, Asa Cooper Stickland, Tomasz Korbak, and Owain Evans. The reversal curse: Llms trained on" a is b" fail to learn" b is a". arXiv preprint arXiv:2309.12288, 2023. find
  16. Jacob Austin, Daniel D Johnson, Jonathan Ho, Daniel Tarlow, and Rianne Van Den Berg. Structured denoising diffusion models in discrete state-spaces. Advances in Neural Information Processing Systems, 34:17981–17993, 2021. find our card
  17. Aaron Lou, Chenlin Meng, and Stefano Ermon. Discrete diffusion language modeling by estimating the ratios of the data distribution. arXiv preprint arXiv:2310.16834, 2023. find
  18. Jiaxin Shi, Kehang Han, Zhe Wang, Arnaud Doucet, and Michalis K Titsias. Simplified and generalized masked diffusion for discrete data. arXiv preprint arXiv:2406.04329, 2024. find
  19. Subham Sekhar Sahoo, Marianne Arriola, Yair Schiff, Aaron Gokaslan, Edgar Marroquin, Justin T Chiu, Alexander Rush, and Volodymyr Kuleshov. Simple and effective masked diffusion language models. arXiv preprint arXiv:2406.07524, 2024. find
  20. Jingyang Ou, Shen Nie, Kaiwen Xue, Fengqi Zhu, Jiacheng Sun, Zhenguo Li, and Chongxuan Li. Your absorbing discrete diffusion secretly models the conditional distributions of clean data. arXiv preprint arXiv:2406.03736, 2024. — [arXiv:2406.03736](https://arxiv.org/abs/2406.03736) find
  21. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288, 2023. find
  22. Jacob Devlin. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018. find
  23. Huiwen Chang, Han Zhang, Lu Jiang, Ce Liu, and William T Freeman. Maskgit: Masked generative image transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11315–11325, 2022. find
  24. Joshua Ainslie, James Lee-Thorp, Michiel de Jong, Yury Zemlyanskiy, Federico Lebron, and Sumit Sanghai. Gqa: Training generalized multi-query transformer models from multi-head checkpoints. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 4895–4901, 20… find our card
  25. An Yang, Baosong Yang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Zhou, Chengpeng Li, Chengyuan Li, Dayiheng Liu, Fei Huang, Guanting Dong, Haoran Wei, Huan Lin, Jialong Tang, Jialin Wang, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Ma, Jianxin Yang, Jin Xu, Jingren Zhou, Jinze Bai, Jinzheng He, Juny… find
  26. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, Keming Lu, Keqin Bao, Kexin Yang, Le Yu, Mei Li, Mingfeng Xue, Pei… find
  27. Shen Nie, Fengqi Zhu, Chao Du, Tianyu Pang, Qian Liu, Guangtao Zeng, Min Lin, and Chongxuan Li. Scaling up masked diffusion models on text. arXiv preprint arXiv:2410.18514, 2024. — [arXiv:2410.18514](https://arxiv.org/abs/2410.18514) find
  28. Shengding Hu, Yuge Tu, Xu Han, Chaoqun He, Ganqu Cui, Xiang Long, Zhi Zheng, Yewei Fang, Yuxiang Huang, Weilin Zhao, et al. Minicpm: Unveiling the potential of small language models with scalable training strategies. arXiv preprint arXiv:2404.06395, 2024. find our card
  29. I Loshchilov. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017. find
  30. Ari Holtzman, Jan Buys, Li Du, Maxwell Forbes, and Yejin Choi. The curious case of neural text degeneration. arXiv preprint arXiv:1904.09751, 2019. find
  31. Marianne Arriola, Aaron Gokaslan, Justin T Chiu, Zhihan Yang, Zhixuan Qi, Jiaqi Han, Subham Sekhar Sahoo, and Volodymyr Kuleshov. Block diffusion: Interpolating between autoregressive and diffusion language models. arXiv preprint arXiv:2503.09573, 2025. — [arXiv:2503.09573](https://arxiv.org/abs/2503.09573) find
  32. Xiao Bi, Deli Chen, Guanting Chen, Shanhuang Chen, Damai Dai, Chengqi Deng, Honghui Ding, Kai Dong, Qiushi Du, Zhe Fu, et al. Deepseek llm: Scaling open-source language models with longtermism. arXiv preprint arXiv:2401.02954, 2024. find
  33. Albert Q Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, et al. Mistral 7b. arXiv preprint arXiv:2310.06825, 2023. find
  34. Tian Ye, Zicheng Xu, Yuanzhi Li, and Zeyuan Allen-Zhu. Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process. ArXiv e-prints, abs/2407.20311, July 2024. Full version available at http://arxiv.org/abs/2407.20311. ↗
  35. Zeyuan Allen-Zhu and Yuanzhi Li. Physics of Language Models: Part 3.2, Knowledge Manipulation. ArXiv e-prints, abs/2309.14402, September 2023. Full version available at http://arxiv.org/abs/2309.14402. ↗
  36. Ouail Kitouni, Niklas Nolte, Diane Bouchacourt, Adina Williams, Mike Rabbat, and Mark Ibrahim. The factorization curse: Which tokens you predict underlie the reversal curse and more. arXiv preprint arXiv:2406.05183, 2024. — [arXiv:2406.05183](https://arxiv.org/abs/2406.05183) find
  37. Jonathan Ho and Tim Salimans. arXiv:2207.12598, 2022. Classifier-free diffusion guidance. arXiv preprint find our card
  38. Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In International conference on machine learning, pages 2256–2265. PMLR, 2015. find our card
  39. Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. find
  40. Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456, 2020. find our card
  41. Xiang Li, John Thickstun, Ishaan Gulrajani, Percy S Liang, and Tatsunori B Hashimoto. Diffusion-lm improves controllable text generation. Advances in Neural Information Processing Systems, 35:4328–4343, 2022. find our card
  42. Shansan Gong, Mukai Li, Jiangtao Feng, Zhiyong Wu, and LingPeng Kong. Diffuseq: Sequence to sequence text generation with diffusion models. arXiv preprint arXiv:2210.08933, 2022. find
  43. Xiaochuang Han, Sachin Kumar, and Yulia Tsvetkov. Ssd-lm: Semi-autoregressive simplex-based diffusion language model for text generation and modular control. arXiv preprint arXiv:2210.17432, 2022. find
  44. Robin Strudel, Corentin Tallec, Florent Altché, Yilun Du, Yaroslav Ganin, Arthur Mensch, Will Grathwohl, Nikolay Savinov, Sander Dieleman, Laurent Sifre, et al. Self-conditioned embedding diffusion for text generation. arXiv preprint arXiv:2211.04236, 2022. — [arXiv:2211.04236](https://arxiv.org/abs/2211.04236) find
  45. Ting Chen, Ruixiang Zhang, and Geoffrey Hinton. Analog bits: Generating discrete data using diffusion models with self-conditioning. arXiv preprint arXiv:2208.04202, 2022. — [arXiv:2208.04202](https://arxiv.org/abs/2208.04202) find
  46. Sander Dieleman, Laurent Sartran, Arman Roshannai, Nikolay Savinov, Yaroslav Ganin, Pierre H Richemond, Arnaud Doucet, Robin Strudel, Chris Dyer, Conor Durkan, et al. Continuous diffusion for categorical data. arXiv preprint arXiv:2211.15089, 2022. — [arXiv:2211.15089](https://arxiv.org/abs/2211.15089) find
  47. Pierre H. Richemond, Sander Dieleman, and Arnaud Doucet. Categorical sdes with simplex diffusion, 2022. find
  48. Tong Wu, Zhihao Fan, Xiao Liu, Yeyun Gong, Yelong Shen, Jian Jiao, Hai-Tao Zheng, Juntao Li, Zhongyu Wei, Jian Guo, Nan Duan, and Weizhu Chen. Ar-diffusion: Auto-regressive diffusion model for text generation, 2023. find
  49. Rabeeh Karimi Mahabadi, Hamish Ivison, Jaesung Tae, James Henderson, Iz Beltagy, Matthew E. Peters, and Arman Cohan. Tess: Text-to-text self-conditioned simplex diffusion, 2024. find
  50. Jiasheng Ye, Zaixiang Zheng, Yu Bao, Lihua Qian, and Mingxuan Wang. Dinoiser: Diffused conditional sequence learning by manipulating noises. arXiv preprint arXiv:2302.10025, 2023. — [arXiv:2302.10025](https://arxiv.org/abs/2302.10025) find
  51. Yizhe Zhang, Jiatao Gu, Zhuofeng Wu, Shuangfei Zhai, Joshua Susskind, and Navdeep Jaitly. Planner: Generating diversified paragraph via latent language diffusion model. Advances in Neural Information Processing Systems, 36:80178–80190, 2023. find
  52. Aaron Lou and Stefano Ermon. Reflected diffusion models, 2023. find
  53. Alex Graves, Rupesh Kumar Srivastava, Timothy Atkinson, and Faustino Gomez. Bayesian flow networks. arXiv preprint arXiv:2308.07037, 2023. — [arXiv:2308.07037](https://arxiv.org/abs/2308.07037) find
  54. Zhenghao Lin, Yeyun Gong, Yelong Shen, Tong Wu, Zhihao Fan, Chen Lin, Nan Duan, and Weizhu Chen. Text generation with diffusion language models: A pre-training approach with continuous paragraph denoise. In International Conference on Machine Learning, pages 21051–21064. PMLR, 2023. find
  55. Kaiwen Xue, Yuhao Zhou, Shen Nie, Xu Min, Xiaolu Zhang, Jun Zhou, and Chongxuan Li. Unifying bayesian flow networks and diffusion models through stochastic differential equations. arXiv preprint arXiv:2404.15766, 2024. — [arXiv:2404.15766](https://arxiv.org/abs/2404.15766) find
  56. Ruixiang Zhang, Shuangfei Zhai, Yizhe Zhang, James Thornton, Zijing Ou, Joshua Susskind, and Navdeep Jaitly. Target concrete score matching: A holistic framework for discrete diffusion. arXiv preprint arXiv:2504.16431, 2025. — [arXiv:2504.16431](https://arxiv.org/abs/2504.16431) find
  57. Ishaan Gulrajani and Tatsunori B Hashimoto. Likelihood-based diffusion language models. Advances in Neural Information Processing Systems, 36, 2024. find
  58. Emiel Hoogeboom, Didrik Nielsen, Priyank Jaini, Patrick Forré, and Max Welling. Argmax flows and multinomial diffusion: Learning categorical distributions. Advances in Neural Information Processing Systems, 34:12454–12465, 2021. find
  59. Emiel Hoogeboom, Alexey A Gritsenko, Jasmijn Bastings, Ben Poole, Rianne van den Berg, and Tim Salimans. Autoregressive diffusion models. arXiv preprint arXiv:2110.02037, 2021. — [arXiv:2110.02037](https://arxiv.org/abs/2110.02037) find
  60. Zhengfu He, Tianxiang Sun, Kuanning Wang, Xuanjing Huang, and Xipeng Qiu. Diffusionbert: Improving generative masked language models with diffusion models. arXiv preprint arXiv:2211.15029, 2022. find
  61. Andrew Campbell, Joe Benton, Valentin De Bortoli, Thomas Rainforth, George Deligiannidis, and Arnaud Doucet. A continuous time framework for discrete denoising models. Advances in Neural Information Processing Systems, 35:28266–28279, 2022. find
  62. Chenlin Meng, Kristy Choi, Jiaming Song, and Stefano Ermon. Concrete score matching: Generalized score matching for discrete data. Advances in Neural Information Processing Systems, 35:34532–34545, 2022. find
  63. Machel Reid, Vincent J. Hellendoorn, and Graham Neubig. Diffuser: Discrete diffusion via edit-based reconstruction, 2022. find
  64. Haoran Sun, Lijun Yu, Bo Dai, Dale Schuurmans, and Hanjun Dai. Score-based continuous-time discrete diffusion models. arXiv preprint arXiv:2211.16750, 2022. — [arXiv:2211.16750](https://arxiv.org/abs/2211.16750) find
  65. Ouail Kitouni, Niklas Nolte, James Hensman, and Bhaskar Mitra. Disk: A diffusion model for structured knowledge. arXiv preprint arXiv:2312.05253, 2023. — [arXiv:2312.05253](https://arxiv.org/abs/2312.05253) find
  66. Lin Zheng, Jianbo Yuan, Lei Yu, and Lingpeng Kong. A reparameterized discrete diffusion model for text generation. ArXiv, abs/2302.05737, 2023. — [arXiv:2302.05737](https://arxiv.org/abs/2302.05737) find
  67. Zixiang Chen, Huizhuo Yuan, Yongqian Li, Yiwen Kou, Junkai Zhang, and Quanquan Gu. Fast sampling via de-randomization for discrete diffusion models. arXiv preprint arXiv:2312.09193, 2023. — [arXiv:2312.09193](https://arxiv.org/abs/2312.09193) find
  68. Jiasheng Ye, Zaixiang Zheng, Yu Bao, Lihua Qian, and Quanquan Gu. Diffusion language models can perform many tasks with scaling and instruction-finetuning. arXiv preprint arXiv:2308.12219, 2023. — [arXiv:2308.12219](https://arxiv.org/abs/2308.12219) find
  69. Itai Gat, Tal Remez, Neta Shaul, Felix Kreuk, Ricky TQ Chen, Gabriel Synnaeve, Yossi Adi, and Yaron Lipman. Discrete flow matching. arXiv preprint arXiv:2407.15595, 2024. — [arXiv:2407.15595](https://arxiv.org/abs/2407.15595) find
  70. Kaiwen Zheng, Yongxin Chen, Hanzi Mao, Ming-Yu Liu, Jun Zhu, and Qinsheng Zhang. Masked diffusion models are secretly time-agnostic masked models and exploit inaccurate categorical sampling, 2024. URL https://arxiv.org/abs/2409.02908. — [arXiv:2409.02908](https://arxiv.org/abs/2409.02908) ↗
  71. Shreyas Kapur, Erik Jenner, and Stuart Russell. Diffusion on syntax trees for program synthesis. arXiv preprint arXiv:2405.20519, 2024. — [arXiv:2405.20519](https://arxiv.org/abs/2405.20519) find
  72. Shansan Gong, Shivam Agarwal, Yizhe Zhang, Jiacheng Ye, Lin Zheng, Mukai Li, Chenxin An, Peilin Zhao, Wei Bi, Jiawei Han, et al. Scaling diffusion language models via adaptation from autoregressive models. arXiv preprint arXiv:2410.17891, 2024. — [arXiv:2410.17891](https://arxiv.org/abs/2410.17891) find
  73. Samar Khanna, Siddhant Kharbanda, Shufan Li, Harshit Varma, Eric Wang, Sawyer Birnbaum, Ziyang Luo, Yanis Miraoui, Akash Palrecha, Stefano Ermon, et al. Mercury: Ultra-fast language models based on diffusion. arXiv preprint arXiv:2506.17298, 2025. find
  74. Huiwen Chang, Han Zhang, Jarred Barber, AJ Maschinot, Jose Lezama, Lu Jiang, Ming-Hsuan Yang, Kevin Murphy, William T Freeman, Michael Rubinstein, et al. Muse: Text-to-image generation via masked generative transformers. arXiv preprint arXiv:2301.00704, 2023. — [arXiv:2301.00704](https://arxiv.org/abs/2301.00704) find
  75. Zebin You, Jingyang Ou, Xiaolu Zhang, Jun Hu, Jun Zhou, and Chongxuan Li. Effective and efficient masked image generation models. arXiv preprint arXiv:2503.07197, 2025. — [arXiv:2503.07197](https://arxiv.org/abs/2503.07197) find
  76. Xinyou Wang, Zaixiang Zheng, Fei Ye, Dongyu Xue, Shujian Huang, and Quanquan Gu. Diffusion language models are versatile protein learners. arXiv preprint arXiv:2402.18567, 2024. — [arXiv:2402.18567](https://arxiv.org/abs/2402.18567) find
  77. Xinyou Wang, Zaixiang Zheng, Fei Ye, Dongyu Xue, Shujian Huang, and Quanquan Gu. Dplm-2: A multimodal diffusion protein language model. arXiv preprint arXiv:2410.13782, 2024. — [arXiv:2410.13782](https://arxiv.org/abs/2410.13782) find
  78. Siqi Kou, Lanxiang Hu, Zhezhi He, Zhijie Deng, and Hao Zhang. Cllms: Consistency large language models. arXiv preprint arXiv:2403.00835, 2024. — [arXiv:2403.00835](https://arxiv.org/abs/2403.00835) find
  79. Chenkai Xu, Xu Wang, Zhenyi Liao, Yishun Li, Tianqi Hou, and Zhijie Deng. Show-o turbo: Towards accelerated unified multimodal understanding and generation. arXiv preprint arXiv:2502.05415, 2025. — [arXiv:2502.05415](https://arxiv.org/abs/2502.05415) find
  80. Sulin Liu, Juno Nam, Andrew Campbell, Hannes Stärk, Yilun Xu, Tommi Jaakkola, and Rafael Gómez-Bombarelli. Think while you generate: Discrete diffusion with planned denoising. arXiv preprint arXiv:2410.06264, 2024. — [arXiv:2410.06264](https://arxiv.org/abs/2410.06264) find
  81. Yuanzhi Zhu, Xi Wang, Stéphane Lathuilière, and Vicky Kalogeiton. Dimo: Distilling masked diffusion models into one-step generator. arXiv preprint arXiv:2503.15457, 2025. — [arXiv:2503.15457](https://arxiv.org/abs/2503.15457) find
  82. Yinuo Ren, Haoxuan Chen, Yuchen Zhu, Wei Guo, Yongxin Chen, Grant M Rotskoff, Molei Tao, and Lexing Ying. Fast solvers for discrete diffusion models: Theory and applications of high-order algorithms. arXiv preprint arXiv:2502.00234, 2025. — [arXiv:2502.00234](https://arxiv.org/abs/2502.00234) find
  83. Satoshi Hayakawa, Yuhta Takida, Masaaki Imaizumi, Hiromi Wakaki, and Yuki Mitsufuji. Distillation of discrete diffusion through dimensional correlations. arXiv preprint arXiv:2410.08709, 2024. — [arXiv:2410.08709](https://arxiv.org/abs/2410.08709) find
  84. Yixiu Zhao, Jiaxin Shi, Feng Chen, Shaul Druckmann, Lester Mackey, and Scott Linderman. Informed correctors for discrete diffusion models. arXiv preprint arXiv:2407.21243, 2024. — [arXiv:2407.21243](https://arxiv.org/abs/2407.21243) find
  85. Kaiwen Zheng, Yongxin Chen, Hanzi Mao, Ming-Yu Liu, Jun Zhu, and Qinsheng Zhang. Masked diffusion models are secretly time-agnostic masked models and exploit inaccurate categorical sampling. arXiv preprint arXiv:2409.02908, 2024. — [arXiv:2409.02908](https://arxiv.org/abs/2409.02908) find
  86. Yong-Hyun Park, Chieh-Hsin Lai, Satoshi Hayakawa, Yuhta Takida, and Yuki Mitsufuji. Jump your steps: Optimizing sampling schedule of discrete diffusion models. In The Thirteenth International Conference on Learning Representations, 2024. find
  87. Justin Deschenaux and Caglar Gulcehre. Beyond autoregression: Fast llms via self-distillation through time. arXiv preprint arXiv:2410.21035, 2024. find
  88. Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural information processing systems, 34:8780–8794, 2021. find
  89. Yair Schiff, Subham Sekhar Sahoo, Hao Phung, Guanghan Wang, Sam Boshar, Hugo Dalla-torre, Bernardo P de Almeida, Alexander Rush, Thomas Pierrot, and Volodymyr Kuleshov. Simple guidance mechanisms for discrete diffusion models. arXiv preprint arXiv:2412.10193, 2024. — [arXiv:2412.10193](https://arxiv.org/abs/2412.10193) find
  90. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 35:27730–27744, 2022. find
  91. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. Direct preference optimization: Your language model is secretly a reward model. Advances in Neural Information Processing Systems, 36, 2024. find
  92. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023. find our card
  93. Google. Our next-generation model: Gemini 1.5, 2024. URL https://blog.google/technology/ai/google-gemini-next-generation-model-february-2024. ↗
  94. Anthropic. Claude 3.5 sonnet, 2024. claude-3-5-sonnet. find
  95. Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, et al. Deepseek-v3 technical report. arXiv preprint arXiv:2412.19437, 2024. find
  96. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35:24824–24837, 2022. find our card
  97. Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th annual acm symposium on user interface software and technology, pages 1–22, 2023. find
  98. Lei Wang, Chen Ma, Xueyang Feng, Zeyu Zhang, Hao Yang, Jingsen Zhang, Zhiyuan Chen, Jiakai Tang, Xu Chen, Yankai Lin, et al. A survey on large language model based autonomous agents. Frontiers of Computer Science, 18(6):186345, 2024. find our card
  99. OpenAI. Learning to reason with llms, 2024. learning-to-reason-with-llms/. find
  100. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948, 2025. find our card
  101. Benigno Uria, Iain Murray, and Hugo Larochelle. A deep and tractable density estimator. In Proceedings of the 31th International Conference on Machine Learning, 2014. find
  102. Andy Shih, Dorsa Sadigh, and Stefano Ermon. Training and inference on any-order autoregressive models the right way. In Proceedings of the 31th International Conference on Machine Learning, 2022. find
  103. Zhangchen Xu, Fengqing Jiang, Luyao Niu, Yuntian Deng, Radha Poovendran, Yejin Choi, and Bill Yuchen Lin. Magpie: Alignment data synthesis from scratch by prompting aligned llms with nothing. arXiv preprint arXiv:2406.08464, 2024. find our card
  104. Yuxiang Wei, Zhe Wang, Jiawei Liu, Yifeng Ding, and Lingming Zhang. Magicoder: Empowering code generation with oss-instruct. arXiv preprint arXiv:2312.02120, 2023. find
  105. Biao Zhang and Rico Sennrich. Root mean square layer normalization. Advances in Neural Information Processing Systems, 32, 2019. find
  106. Noam Shazeer. Glu variants improve transformer. arXiv preprint arXiv:2002.05202, 2020. find
  107. Jianlin Su, Murtadha Ahmed, Yu Lu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. Roformer: Enhanced transformer with rotary position embedding. Neurocomputing, 568:127063, 2024. find
  108. Jared Kaplan, Sam McCandlish, Tom Henighan, Tom B Brown, Benjamin Chess, Rewon Child, Scott Gray, Alec Radford, Jeffrey Wu, and Dario Amodei. Scaling laws for neural language models. arXiv preprint arXiv:2001.08361, 2020. URL https://www.anthropic.com/news/URL https://openai.com/index/ ↗
  109. Jordan Hoffmann, Sebastian Borgeaud, Arthur Mensch, Elena Buchatskaya, Trevor Cai, Eliza Rutherford, Diego de Las Casas, Lisa Anne Hendricks, Johannes Welbl, Aidan Clark, et al. Training compute-optimal large language models. arXiv preprint arXiv:2203.15556, 2022. find
  110. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. Measuring massive multitask language understanding. arXiv preprint arXiv:2009.03300, 2020. find our card
  111. Mirac Suzgun, Nathan Scales, Nathanael Schärli, Sebastian Gehrmann, Yi Tay, Hyung Won Chung, Aakanksha Chowdhery, Quoc V Le, Ed H Chi, Denny Zhou, et al. Challenging big-bench tasks and whether chain-of-thought can solve them. arXiv preprint arXiv:2210.09261, 2022. find our card
  112. Peter Clark, Isaac Cowhey, Oren Etzioni, Tushar Khot, Ashish Sabharwal, Carissa Schoenick, and Oyvind Tafjord. Think you have solved question answering? try arc, the ai2 reasoning challenge. arXiv preprint arXiv:1803.05457, 2018. find
  113. Rowan Zellers, Ari Holtzman, Yonatan Bisk, Ali Farhadi, and Yejin Choi. Hellaswag: Can a machine really finish your sentence? arXiv preprint arXiv:1905.07830, 2019. find
  114. Stephanie Lin, Jacob Hilton, and Owain Evans. Truthfulqa: Measuring how models mimic human falsehoods. arXiv preprint arXiv:2109.07958, 2021. find our card
  115. Keisuke Sakaguchi, Ronan Le Bras, Chandra Bhagavatula, and Yejin Choi. Winogrande: An adversarial winograd schema challenge at scale. Communications of the ACM, 64(9):99–106, 2021. find
  116. Yonatan Bisk, Rowan Zellers, Jianfeng Gao, Yejin Choi, et al. Piqa: Reasoning about physical commonsense in natural language. In Proceedings of the AAAI conference on artificial intelligence, 2020. find our card
  117. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, et al. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168, 2021. find our card
  118. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. Measuring mathematical problem solving with the math dataset. arXiv preprint arXiv:2103.03874, 2021. find our card
  119. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R Bowman. Gpqa: A graduate-level google-proof q&a benchmark. arXiv preprint arXiv:2311.12022, 2023. find our card
  120. Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde De Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, et al. Evaluating large language models trained on code. arXiv preprint arXiv:2107.03374, 2021. find our card
  121. Mohammad Bavarian, Heewoo Jun, Nikolas Tezak, John Schulman, Christine McLeavey, Jerry Tworek, and Mark Chen. Efficient training of language models to fill in the middle. arXiv preprint arXiv:2207.14255, 2022. find our card
  122. Jacob Austin, Augustus Odena, Maxwell Nye, Maarten Bosma, Henryk Michalewski, David Dohan, Ellen Jiang, Carrie Cai, Michael Terry, Quoc Le, et al. Program synthesis with large language models. arXiv preprint arXiv:2108.07732, 2021. find our card
  123. Haonan Li, Yixuan Zhang, Fajri Koto, Yifei Yang, Hai Zhao, Yeyun Gong, Nan Duan, and Timothy Baldwin. Cmmlu: Measuring massive multitask language understanding in chinese. arXiv preprint arXiv:2306.09212, 2023. find our card
  124. Yuzhen Huang, Yuzhuo Bai, Zhihao Zhu, Junlei Zhang, Jinghan Zhang, Tangjun Su, Junteng Liu, Chuancheng Lv, Yikai Zhang, Yao Fu, et al. C-eval: A multi-level multi-discipline chinese evaluation suite for foundation models. Advances in Neural Information Processing Systems, 36, 2024. find our card
  125. Leo Gao, Jonathan Tow, Baber Abbasi, Stella Biderman, Sid Black, Anthony DiPofi, Charles Foster, Laurence Golding, Jeffrey Hsu, Alain Le Noac’h, Haonan Li, Kyle McDonell, Niklas Muennighoff, Chris Ociepa, Jason Phang, Laria Reynolds, Hailey Schoelkopf, Aviya Skowron, Lintang Sutawika, Eric Tang, Ani… find