Pith. sign in

REVIEW 90 references

DiverseDiT++: Quantifying, Analyzing, and Promoting Representation Diversity in Diffusion Transformers

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2608.03082 v1 pith:6XI2IFKB submitted 2026-08-04 cs.CV

Binglei Li , Mengping Yang , Zhiyu Tan , Xiaomeng Yang , Zhizhong Huang , Junping Zhang , Hao Li This is my paper · ORCID
classification cs.CV
keywords representationditsdiversityacrossblocksrepresentationsdiverseditlearning
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

Recent advances in Diffusion Transformers (DiTs) have enabled remarkable progress in visual synthesis, benefiting from their superior scalability. To facilitate DiTs' capability of capturing meaningful internal representations, recent works such as REPA incorporate external pretrained encoders for representation alignment. However, the underlying mechanisms governing representation learning within DiTs remain poorly understood in the community. To this end, this paper first presents a systematic analysis of the representation dynamics of DiTs via quantifying the diversity of block-wise representations. Specifically, we introduce a novel metric, termed the Weighted Diversity Score (WDS), to measure the representational discrepancies across different blocks. Through extensive investigations on the evolution and influence of internal representations under various settings, we reveal that representation diversity across blocks is a critical factor for effective representation learning in DiTs. More importantly, WDS exhibits a strong correlation with synthesis quality across diverse settings, model scales, and training stages (Pearson's $r=-0.869$ with $\log(\text{FID})$), suggesting its potential as an indicator to reflect model performance and a principled guide for model optimization. Based on this key finding, we propose DiverseDiT++, a novel framework that explicitly promotes diverse representation learning. Concretely, our method incorporates long residual connections to diversify input representations across blocks and a representation diversity loss to encourage blocks to learn distinct features. Extensive experiments on ImageNet $256\times256$ and $512\times512$ demonstrate that our DiverseDiT++ yields consistent performance gains and convergence acceleration when applied to different backbones with various sizes,...

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

90 extracted references · 27 canonical work pages

  1. [1]

    Rep- resentation alignment for generation: Training diffusion transformers is easier than you think,

    S. Yu, S. Kwak, H. Jang, J. Jeong, J. Huang, J. Shin, and S. Xie, “Rep- resentation alignment for generation: Training diffusion transformers is easier than you think,” inInt. Conf. Learn. Represent., 2024. 1, 2, 3, 4, 5, 8, 9, 16, 18

    2024

  2. [2]

    Diffuse and disperse: Image generation with representation regularization,

    R. Wang and K. He, “Diffuse and disperse: Image generation with representation regularization,”arXiv preprint arXiv:2506.09027, 2025. 1, 2, 3, 8, 9, 10, 11, 16, 19, 22

    arXiv 2025

  3. [3]

    Deep unsupervised learning using nonequilibrium thermodynamics,

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequilibrium thermodynamics,” inInt. Conf. Mach. Learn.PMLR, 2015, pp. 2256–2265. 1, 3

    2015

  4. [4]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Adv. Neural Inform. Process. Syst., vol. 33, pp. 6840–6851, 2020. 1, 3, 4

    2020

  5. [5]

    Scalable diffusion models with transformers,

    W. Peebles and S. Xie, “Scalable diffusion models with transformers,” inInt. Conf. Comput. Vis., 2023, pp. 4195–4205. 1, 3, 8, 9, 18

    2023

  6. [6]

    All are worth words: A vit backbone for diffusion models,

    F. Bao, S. Nie, K. Xue, Y . Cao, C. Li, H. Su, and J. Zhu, “All are worth words: A vit backbone for diffusion models,” inIEEE Conf. Comput. Vis. Pattern Recog., 2023, pp. 22 669–22 679. 1, 3

    2023

  7. [7]

    Sana: Efficient high-resolution text-to-image syn- thesis with linear diffusion transformers,

    E. Xie, J. Chen, J. Chen, H. Cai, H. Tang, Y . Lin, Z. Zhang, M. Li, L. Zhu, Y . Luet al., “Sana: Efficient high-resolution text-to-image syn- thesis with linear diffusion transformers,” inInt. Conf. Learn. Represent.,

  8. [8]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” inIEEE Conf. Comput. Vis. Pattern Recog., 2022, pp. 10 684–10 695. 1, 3, 8, 9, 16, 18

    2022

Show all 90 references
  1. [9]

    Cogvideox: Text-to-video diffusion models with an expert transformer,

    Z. Yang, J. Teng, W. Zheng, M. Ding, S. Huang, J. Xu, Y . Yang, W. Hong, X. Zhang, G. Fenget al., “Cogvideox: Text-to-video diffusion models with an expert transformer,” inInt. Conf. Learn. Represent.,

  2. [10]

    Photorealistic video generation with diffusion models,

    A. Gupta, L. Yu, K. Sohn, X. Gu, M. Hahn, F.-F. Li, I. Essa, L. Jiang, and J. Lezama, “Photorealistic video generation with diffusion models,” inEur. Conf. Comput. Vis.Springer, 2024, pp. 393–411. 1

    2024

  3. [11]

    Diffusion based representation learning,

    S. Mittal, K. Abstreiter, S. Bauer, B. Sch ¨olkopf, and A. Mehrjou, “Diffusion based representation learning,” inInt. Conf. Mach. Learn. PMLR, 2023, pp. 24 963–24 982. 1, 3

    2023

  4. [12]

    Deconstructing denoising diffusion models for self-supervised learning,

    X. Chen, Z. Liu, S. Xie, and K. He, “Deconstructing denoising diffusion models for self-supervised learning,” inInt. Conf. Learn. Represent.,

  5. [13]

    Denoising diffusion autoencoders are unified self-supervised learners,

    W. Xiang, H. Yang, D. Huang, and Y . Wang, “Denoising diffusion autoencoders are unified self-supervised learners,” inInt. Conf. Comput. Vis., 2023, pp. 15 802–15 812. 1, 3

    2023

  6. [14]

    REPA-E: Unlocking vae for end-to-end tuning with latent diffusion transformers,

    X. Leng, J. Singh, Y . Hou, Z. Xing, S. Xie, and L. Zheng, “REPA-E: Unlocking vae for end-to-end tuning with latent diffusion transformers,” inInt. Conf. Comput. Vis., 2025, pp. 18 262–18 272. 1, 3, 8, 9, 18

    2025

  7. [15]

    Representation entanglement for generation: Training diffusion transformers is much easier than you think,

    G. Wu, S. Zhang, R. Shi, S. Gao, Z. Chen, L. Wang, Z. Chen, H. Gao, Y . Tang, M.-M. Chenget al., “Representation entanglement for generation: Training diffusion transformers is much easier than you think,”Adv. Neural Inform. Process. Syst., vol. 38, pp. 7714–7743, 2026. 1, 3, 8, 9, 18

    2026

  8. [16]

    No other representation component is needed: Diffusion transformers can provide representation guidance by themselves,

    D. Jiang, M. Wang, L. Li, L. Zhang, H. Wang, W. Wei, G. Dai, Y . Zhang, and J. Wang, “No other representation component is needed: Diffusion transformers can provide representation guidance by themselves,”arXiv preprint arXiv:2505.02831, 2025. 1, 2, 3, 8, 9, 11, 16, 19, 22

    2025

  9. [17]

    Similarity of neural network representations revisited,

    S. Kornblith, M. Norouzi, H. Lee, and G. Hinton, “Similarity of neural network representations revisited,” inInt. Conf. Mach. Learn.PMLR, 2019, pp. 3519–3529. 1, 3, 5, 15, 17, 18

    2019

  10. [18]

    Mean flows for one- step generative modeling,

    Z. Geng, M. Deng, X. Bai, J. Z. Kolter, and K. He, “Mean flows for one- step generative modeling,”Adv. Neural Inform. Process. Syst., vol. 38, pp. 75 460–75 482, 2026. 2, 8, 9, 10, 16, 19

    2026

  11. [19]

    Diversedit: Towards diverse representation learning in diffusion transformers,

    M. Yang, Z. Tan, B. Li, X. Yang, H. Chen, and H. Li, “Diversedit: Towards diverse representation learning in diffusion transformers,” in IEEE Conf. Comput. Vis. Pattern Recog., 2026. 3, 9

    2026

  12. [20]

    Representation learning: A review and new perspectives,

    Y . Bengio, A. Courville, and P. Vincent, “Representation learning: A review and new perspectives,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 35, no. 8, pp. 1798–1828, 2013. 3

    2013

  13. [21]

    Disentangled rep- resentation learning,

    X. Wang, H. Chen, S. Tang, Z. Wu, and W. Zhu, “Disentangled rep- resentation learning,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 46, no. 12, pp. 9677–9696, 2024. 3

    2024

  14. [22]

    Deep diversity- enhanced feature representation of hyperspectral images,

    J. Hou, Z. Zhu, J. Hou, H. Liu, H. Zeng, and D. Meng, “Deep diversity- enhanced feature representation of hyperspectral images,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 46, no. 12, pp. 8123–8138, 2024. 3

    2024

  15. [23]

    Self-supervised visual feature learning with deep neural networks: A survey,

    L. Jing and Y . Tian, “Self-supervised visual feature learning with deep neural networks: A survey,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 43, no. 11, pp. 4037–4058, 2021. 3

    2021

  16. [24]

    Bootstrap your own latent-a new approach to self-supervised learning,

    J.-B. Grill, F. Strub, F. Altch ´e, C. Tallec, P. Richemond, E. Buchatskaya, C. Doersch, B. Avila Pires, Z. Guo, M. Gheshlaghi Azaret al., “Bootstrap your own latent-a new approach to self-supervised learning,” 13 inAdv. Neural Inform. Process. Syst., vol. 33, 2020, pp. 21 271...

    2020

  17. [25]

    Dinov2: Learning robust visual features without supervision,

    M. Oquab, T. Darcet, T. Moutakanni, H. V o, M. Szafraniec, V . Khalidov, P. Fernandez, D. Haziza, F. Massa, A. El-Noubyet al., “Dinov2: Learning robust visual features without supervision,”Transactions on Machine Learning Research Journal, 2024. 3, 5, 8, 16, 17

    2024

  18. [26]

    An empirical study of training self- supervised vision transformers,

    X. Chen, S. Xie, and K. He, “An empirical study of training self- supervised vision transformers,” inInt. Conf. Comput. Vis., 2021, pp. 9640–9649. 3, 5, 8, 17

    2021

  19. [27]

    Auto-encoding variational bayes,

    D. P. Kingma and M. Welling, “Auto-encoding variational bayes,” in Int. Conf. Learn. Represent., 2014. 3

    2014

  20. [28]

    Masked autoencoders are scalable vision learners,

    K. He, X. Chen, S. Xie, Y . Li, P. Doll ´ar, and R. Girshick, “Masked autoencoders are scalable vision learners,” inIEEE Conf. Comput. Vis. Pattern Recog., 2022, pp. 16 000–16 009. 3, 8, 17, 18

    2022

  21. [29]

    Simmim: A simple framework for masked image modeling,

    Z. Xie, Z. Zhang, Y . Cao, Y . Lin, J. Bao, Z. Yao, Q. Dai, and H. Hu, “Simmim: A simple framework for masked image modeling,” inIEEE Conf. Comput. Vis. Pattern Recog., 2022, pp. 9653–9663. 3

    2022

  22. [30]

    Diffusion model as representation learner,

    X. Yang and X. Wang, “Diffusion model as representation learner,” in Int. Conf. Comput. Vis., 2023, pp. 18 938–18 949. 3

    2023

  23. [31]

    Unsupervised representation learning from pre-trained diffusion probabilistic models,

    Z. Zhang, Z. Zhao, and Z. Lin, “Unsupervised representation learning from pre-trained diffusion probabilistic models,” inAdv. Neural Inform. Process. Syst., vol. 35, 2022, pp. 22 117–22 130. 3

    2022

  24. [32]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clarket al., “Learning transferable visual models from natural language supervision,” inInt. Conf. Mach. Learn., 2021, pp. 8748–8763. 3

    2021

  25. [33]

    Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation,

    J. Li, D. Li, C. Xiong, and S. Hoi, “Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation,” inInt. Conf. Mach. Learn.PMLR, 2022, pp. 12 888–12 900. 3

    2022

  26. [34]

    Sigmoid loss for language image pre-training,

    X. Zhai, B. Mustafa, A. Kolesnikov, and L. Beyer, “Sigmoid loss for language image pre-training,” inInt. Conf. Comput. Vis., 2023, pp. 11 975–11 986. 3

    2023

  27. [35]

    Siglip 2: Multilingual vision-language encoders with improved semantic understanding, localization, and dense features,

    M. Tschannen, A. Gritsenko, X. Wang, M. F. Naeem, I. Alabdul- mohsin, N. Parthasarathy, T. Evans, L. Beyer, Y . Xia, B. Mustafa et al., “Siglip 2: Multilingual vision-language encoders with improved semantic understanding, localization, and dense features,”arXiv preprint arXiv...

    2025 arXiv

  28. [36]

    ImageNet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “ImageNet: A large-scale hierarchical image database,” inIEEE Conf. Comput. Vis. Pattern Recog., 2009, pp. 248–255. 3

    2009

  29. [37]

    Assessing generative models via precision and recall,

    M. S. Sajjadi, O. Bachem, M. Lucic, O. Bousquet, and S. Gelly, “Assessing generative models via precision and recall,”Adv. Neural Inform. Process. Syst., vol. 31, 2018. 3

    2018

  30. [38]

    Im- proved precision and recall metric for assessing generative models,

    T. Kynk ¨a¨anniemi, T. Karras, S. Laine, J. Lehtinen, and T. Aila, “Im- proved precision and recall metric for assessing generative models,” in Adv. Neural Inform. Process. Syst., vol. 32, 2019. 3, 8, 17

    2019

  31. [39]

    Reliability of cka as a similarity measure in deep learning,

    M. Davari, S. Horoi, A. Natik, G. Lajoie, G. Wolf, and E. Belilovsky, “Reliability of cka as a similarity measure in deep learning,” inInt. Conf. Learn. Represent., 2023. 3, 5, 17

    2023

  32. [40]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium,

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,”Adv. Neural Inform. Process. Syst., vol. 30, 2017. 3, 6, 8, 15, 17

    2017

  33. [41]

    Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models,

    G. Stein, J. Cresswell, R. Hosseinzadeh, Y . Sui, B. Ross, V . Villecroze, Z. Liu, A. L. Caterini, E. Taylor, and G. Loaiza-Ganem, “Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models,”Adv. Neural Inform. Process. Syst., vol. 36...

    2023

  34. [42]

    Generating images with sparse representations,

    C. Nash, J. Menick, S. Dieleman, and P. Battaglia, “Generating images with sparse representations,” inInt. Conf. Mach. Learn.PMLR, 2021, pp. 7958–7968. 3, 8, 17

    2021

  35. [43]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” inInt. Conf. Learn. Represent., 2021. 3

    2021

  36. [44]

    Diffusion models in vision: A survey,

    F.-A. Croitoru, V . Hondru, R. T. Ionescu, and M. Shah, “Diffusion models in vision: A survey,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 45, no. 9, pp. 10 850–10 869, 2023. 3

    2023

  37. [45]

    Pixart-α: Fast training of diffusion transformer for photorealistic text-to-image synthesis,

    J. Chen, J. Yu, C. Ge, L. Yao, E. Xie, Y . Wu, Z. Wang, J. Kwok, P. Luo, H. Luet al., “Pixart-α: Fast training of diffusion transformer for photorealistic text-to-image synthesis,” inInt. Conf. Learn. Represent.,

  38. [46]

    Hunyuanvideo: A systematic framework for large video generative models,

    W. Kong, Q. Tian, Z. Zhang, R. Min, Z. Dai, J. Zhou, J. Xiong, X. Li, B. Wu, J. Zhanget al., “Hunyuanvideo: A systematic framework for large video generative models,”arXiv preprint arXiv:2412.03603, 2024. 3

    2024 arXiv

  39. [47]

    Flow matching for generative modeling,

    Y . Lipman, R. T. Chen, H. Ben-Hamu, M. Nickel, and M. Le, “Flow matching for generative modeling,” inInt. Conf. Learn. Represent., 2023. 3, 4

    2023

  40. [48]

    Flow straight and fast: Learning to generate and transfer data with rectified flow,

    X. Liu, C. Gonget al., “Flow straight and fast: Learning to generate and transfer data with rectified flow,” inInt. Conf. Learn. Represent.,

  41. [49]

    Diffusion models beat gans on image synthesis,

    P. Dhariwal and A. Nichol, “Diffusion models beat gans on image synthesis,” inAdv. Neural Inform. Process. Syst., vol. 34, 2021, pp. 8780–8794. 3, 8, 9, 17, 18

    2021

  42. [50]

    Stable video diffusion: Scaling latent video diffusion models to large datasets,

    A. Blattmann, T. Dockhorn, S. Kulal, D. Mendelevitch, M. Kilian, D. Lorenz, Y . Levi, Z. English, V . V oleti, A. Lettset al., “Stable video diffusion: Scaling latent video diffusion models to large datasets,”arXiv preprint arXiv:2311.15127, 2023. 3

    2023 arXiv

  43. [51]

    SiT: Exploring flow and diffusion-based generative models with scalable interpolant transformers,

    N. Ma, M. Goldstein, M. S. Albergo, N. M. Boffi, E. Vanden-Eijnden, and S. Xie, “SiT: Exploring flow and diffusion-based generative models with scalable interpolant transformers,” inEur. Conf. Comput. Vis. Springer, 2024, pp. 23–40. 3, 8, 9, 16, 18

    2024

  44. [52]

    Dreamteacher: Pretraining image backbones with deep generative models,

    D. Li, H. Ling, A. Kar, D. Acuna, S. W. Kim, K. Kreis, A. Torralba, and S. Fidler, “Dreamteacher: Pretraining image backbones with deep generative models,” inInt. Conf. Comput. Vis., 2023, pp. 16 698–16 708. 3

    2023

  45. [53]

    Diffusion models and representation learning: A survey,

    M. Fuest, P. Ma, M. Gui, J. Schusterbauer, V . T. Hu, and B. Ommer, “Diffusion models and representation learning: A survey,”IEEE Trans. Pattern Anal. Mach. Intell., 2026. 3

    2026

  46. [54]

    SARA: Structural and adversarial representation alignment for training-efficient diffusion models,

    H. Chen, J. Wang, Z. Tan, and H. Li, “SARA: Structural and adversarial representation alignment for training-efficient diffusion models,”arXiv preprint arXiv:2503.08253, 2025. 3

    2025 arXiv

  47. [55]

    Aligning text to image in diffusion models is easier than you think,

    J.-Y . Lee, B. Cha, J. Kim, and J. C. Ye, “Aligning text to image in diffusion models is easier than you think,”Adv. Neural Inform. Process. Syst., vol. 38, pp. 157 106–157 136, 2026. 3

    2026

  48. [56]

    Learning diffusion models with flexible representation guidance,

    C. Wang, C. Zhou, S. Gupta, J. Lin, S. Jegelka, S. Bates, and T. Jaakkola, “Learning diffusion models with flexible representation guidance,”Adv. Neural Inform. Process. Syst., vol. 38, pp. 131 176–131 222, 2026. 3, 4, 10, 19

    2026

  49. [57]

    What matters for representation alignment: Global information or spatial structure?

    J. Singh, X. Leng, Z. Wu, L. Zheng, R. Zhang, E. Shechtman, and S. Xie, “What matters for representation alignment: Global information or spatial structure?” inInt. Conf. Learn. Represent., 2026. 3, 9

    2026

  50. [58]

    Generative flows on discrete state-spaces: enabling multimodal flows with applications to protein co-design,

    A. Campbell, J. Yim, R. Barzilay, T. Rainforth, and T. Jaakkola, “Generative flows on discrete state-spaces: enabling multimodal flows with applications to protein co-design,” inInt. Conf. Mach. Learn., 2024, pp. 5453–5512. 3, 10, 16, 18, 19

    2024

  51. [59]

    Ro- bust deep learning–based protein sequence design using proteinmpnn,

    J. Dauparas, I. Anishchenko, N. Bennett, H. Bai, R. J. Ragotte, L. F. Milles, B. I. Wicky, A. Courbet, R. J. de Haas, N. Bethelet al., “Ro- bust deep learning–based protein sequence design using proteinmpnn,” Science, vol. 378, no. 6615, pp. 49–56, 2022. 3, 10, 16, 19

    2022

  52. [60]

    MiDi: Mixed graph and 3d denoising diffusion for molecule generation,

    C. Vignac, N. Osman, L. Toni, and P. Frossard, “MiDi: Mixed graph and 3d denoising diffusion for molecule generation,” inJoint European Con- ference on Machine Learning and Knowledge Discovery in Databases, 2023, pp. 560–576. 4

    2023

  53. [61]

    Navigating the design space of equivariant diffusion-based generative models for de novo 3d molecule generation,

    T. Le, J. Cremer, F. Noe, D.-A. Clevert, and K. T. Sch ¨utt, “Navigating the design space of equivariant diffusion-based generative models for de novo 3d molecule generation,” inInt. Conf. Learn. Represent., 2024. 4

    2024

  54. [62]

    Semlaflow–efficient 3d molecular generation with latent attention and equivariant flow matching,

    R. Irwin, A. Tibo, J. P. Janet, and S. Olsson, “Semlaflow–efficient 3d molecular generation with latent attention and equivariant flow matching,”arXiv preprint arXiv:2406.07266, 2024. 4, 11, 17, 19

    2024 arXiv

  55. [63]

    Accurate structure prediction of biomolecular interactions with AlphaFold 3,

    J. Abramson, J. Adler, J. Dunber, R. Evanset al., “Accurate structure prediction of biomolecular interactions with AlphaFold 3,”Nature, vol. 630, pp. 493–500, 2024. 4, 19

    2024

  56. [64]

    Uni-mol: A universal 3d molecular representation learning framework,

    G. Zhou, Z. Gao, Q. Ding, H. Zheng, H. Xu, Z. Wei, L. Zhang, and G. Ke, “Uni-mol: A universal 3d molecular representation learning framework,” inInt. Conf. Learn. Represent., 2023. 4, 19

    2023

  57. [65]

    Measuring statistical dependence with hilbert-schmidt norms,

    A. Gretton, O. Bousquet, A. Smola, and B. Sch ¨olkopf, “Measuring statistical dependence with hilbert-schmidt norms,” inInternational Conference on Algorithmic Learning Theory, 2005, pp. 63–77. 5, 18

    2005

  58. [66]

    Masked autoencoders are scalable vision learners,

    K. He, X. Chen, S. Xie, Y . Li, P. Doll ´ar, and R. Girshick, “Masked autoencoders are scalable vision learners,” inIEEE Conf. Comput. Vis. Pattern Recog., 2022. 5

    2022

  59. [67]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” inIEEE Conf. Comput. Vis. Pattern Recog., 2009, pp. 248–255. 8

    2009

  60. [68]

    Improved techniques for training gans,

    T. Salimans, I. Goodfellow, W. Zaremba, V . Cheung, A. Radford, and X. Chen, “Improved techniques for training gans,” inAdv. Neural Inform. Process. Syst., vol. 29, 2016. 8, 17

    2016

  61. [69]

    Classifier-free diffusion guidance,

    J. Ho and T. Salimans, “Classifier-free diffusion guidance,”arXiv preprint arXiv:2207.12598, 2022. 8

    2022 arXiv

  62. [70]

    Understanding diffusion objectives as the elbo with simple data augmentation,

    D. Kingma and R. Gao, “Understanding diffusion objectives as the elbo with simple data augmentation,”Adv. Neural Inform. Process. Syst., vol. 36, pp. 65 484–65 516, 2023. 8, 9, 18 14

    2023

  63. [71]

    Cascaded diffusion models for high fidelity image generation,

    J. Ho, C. Saharia, W. Chan, D. J. Fleet, M. Norouzi, and T. Salimans, “Cascaded diffusion models for high fidelity image generation,”Journal of Machine Learning Research, vol. 23, no. 47, pp. 1–33, 2022. 8, 9, 18

    2022

  64. [72]

    Sd- dit: Unleashing the power of self-supervised discrimination in diffusion transformer,

    R. Zhu, Y . Pan, Y . Li, T. Yao, Z. Sun, T. Mei, and C. W. Chen, “Sd- dit: Unleashing the power of self-supervised discrimination in diffusion transformer,” inIEEE Conf. Comput. Vis. Pattern Recog., 2024, pp. 8435–8445. 8, 9, 18

    2024

  65. [73]

    Fast train- ing of diffusion models with masked transformers,

    H. Zheng, W. Nie, A. Vahdat, and A. Anandkumar, “Fast train- ing of diffusion models with masked transformers,”arXiv preprint arXiv:2306.09305, 2023. 8, 9, 18

    2023 arXiv

  66. [74]

    Mdtv2: Masked dif- fusion transformer is a strong image synthesizer,

    S. Gao, P. Zhou, M.-M. Cheng, and S. Yan, “Mdtv2: Masked dif- fusion transformer is a strong image synthesizer,”arXiv preprint arXiv:2303.14389, 2023. 8, 9, 18

    2023 arXiv

  67. [75]

    Improved techniques for training consistency models,

    Y . Song and P. Dhariwal, “Improved techniques for training consistency models,” inInt. Conf. Learn. Represent., 2024. 8, 10, 19

    2024

  68. [76]

    One step diffusion via shortcut models,

    K. Frans, D. Hafner, S. Levine, and P. Abbeel, “One step diffusion via shortcut models,” inInt. Conf. Learn. Represent., 2025. 8, 10, 19

    2025

  69. [77]

    Inductive moment matching,

    L. Zhou, S. Ermon, and J. Song, “Inductive moment matching,” inInt. Conf. Mach. Learn.PMLR, 2025, pp. 78 651–78 686. 8, 10, 19

    2025

  70. [78]

    Fine-tuning discrete diffusion models via reward optimization with applications to dna and protein design,

    C. Wang, M. Uehara, Y . He, A. Wang, A. Lal, T. Jaakkola, S. Levine, A. Regev, H. Wang, and T. Biancalani, “Fine-tuning discrete diffusion models via reward optimization with applications to dna and protein design,” inInt. Conf. Learn. Represent., 2025. 10, 16

    2025

  71. [79]

    Evolutionary-scale prediction of atomic-level protein structure with a language model,

    Z. Lin, H. Akin, R. Rao, B. Hie, Z. Zhuet al., “Evolutionary-scale prediction of atomic-level protein structure with a language model,” Science, vol. 379, no. 6637, pp. 1123–1130, 2023. 10, 18, 23

    2023

  72. [80]

    GEOM, energy-annotated molecular conformations for property prediction and molecular genera- tion,

    S. Axelrod and R. Gomez-Bombarelli, “GEOM, energy-annotated molecular conformations for property prediction and molecular genera- tion,”Scientific data, vol. 9, no. 1, p. 185, 2022. 11, 17, 24

    2022

  73. [81]

    Quantum chemistry structures and properties of 134 kilo molecules,

    R. Ramakrishnan, P. O. Dral, M. Rupp, and O. A. V on Lilienfeld, “Quantum chemistry structures and properties of 134 kilo molecules,” Scientific data, vol. 1, no. 1, pp. 1–7, 2014. 11, 25

    2014

  74. [82]

    The effective rank: A measure of effective dimensionality,

    O. Roy and M. Vetterli, “The effective rank: A measure of effective dimensionality,”European Signal Processing Conference, pp. 606–610,

  75. [83]

    Visualizing and understanding convolu- tional networks,

    M. D. Zeiler and R. Fergus, “Visualizing and understanding convolu- tional networks,” inEur. Conf. Comput. Vis., 2014. 15

    2014

  76. [84]

    Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions,

    S. Chen, S. Chewi, J. Li, Y . Li, A. Salim, and A. R. Zhang, “Sampling is as easy as learning the score: theory for diffusion models with minimal data assumptions,”arXiv preprint arXiv:2209.11215, 2022. 15

    2022 arXiv

  77. [85]

    Decoupled weight decay regularization,

    I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” arXiv preprint arXiv:1711.05101, 2017. 16

    2017 arXiv

  78. [86]

    Return of unconditional generation: A self- supervised representation generation method,

    T. Li, D. Katabi, and K. He, “Return of unconditional generation: A self- supervised representation generation method,” inAdv. Neural Inform. Process. Syst., vol. 37, 2024, pp. 125 441–125 468. 17

    2024

  79. [87]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” inIEEE Conf. Comput. Vis. Pattern Recog., 2020, pp. 9729–9738. 17

    2020

  80. [88]

    Improved baselines with mo- mentum contrastive learning,

    X. Chen, H. Fan, R. Girshick, and K. He, “Improved baselines with mo- mentum contrastive learning,”arXiv preprint arXiv:2003.04297, 2020. 17

    2003 arXiv

  81. [89]

    Rethinking the inception architecture for computer vision,

    C. Szegedy, V . Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethinking the inception architecture for computer vision,” inIEEE Conf. Comput. Vis. Pattern Recog., 2016, pp. 2818–2826. 17

    2016

  82. [90]

    Improved denoising diffusion probabilis- tic models,

    A. Q. Nichol and P. Dhariwal, “Improved denoising diffusion probabilis- tic models,” inInt. Conf. Mach. Learn.PMLR, 2021, pp. 8162–8171. 19 15 APPENDIXA APPENDIXOVERVIEW This supplementary material is organized as follows: First, we provide a heuristic analysis of the WDS–FID ...

    2021

Pith tools