- 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

Machine-Learned Particle Flow as a Foundation Model for Collider Physics

Farouk Mokhtar1,2,*, Joosep Pata3, Michael Kagan4, and Javier Duarte1

  • *Contact author: fmokhtar@ucsd.edu
PDF

PRX Intelligence 1, 023012 – Published 4 November, 2026

DOI: https://doi.org/10.1103/v5n3-d3tn

Abstract

The workflow from particle collision to physics analysis passes through a series of reconstruction steps that are traditionally modular and disconnected, with no shared representation linking low-level detector data to high-level analysis tasks. We show that casting event reconstruction as a machine learning problem naturally produces such a shared representation. We repurpose a machine learning model trained for particle-flow reconstruction (MLPF) to perform three distinct analysis tasks: jet flavor identification, jet energy regression, and missing momentum regression. By appending the per-particle latent representations learned during reconstruction as additional input features, we substantially improve over baselines that use conventional particle-level features alone. We further demonstrate that a single linear layer trained using only the latent representations achieves competitive performance against state-of-the-art baseline architectures, and outperforms the baseline for missing momentum regression with approximately 35 times fewer parameters. These results demonstrate that the latent representations learned during reconstruction encode essential physics information needed for downstream analysis, establishing MLPF as a foundation model and offering a concrete step toward an end-to-end pipeline from detector data to physics analysis.

View figure in article

Physics Subject Headings (PhySH)

Article Text

References (49)

  1. A. M. Sirunyan et al. (CMS Collaboration), Particle-flow reconstruction and global event description with the CMS detector, J. Instrum. 12, P10003 (2017).
  2. D. Buskulic et al. (ALEPH Collaboration), Performance of the ALEPH detector at LEP, Nucl. Instrum. Methods Phys. Res. Sect. A 360, 481 (1995).
  3. J. Pata, J. Duarte, J.-R. Vlimant, M. Pierini, and M. Spiropulu, MLPF: Efficient machine-learned particle-flow reconstruction using graph neural networks, Eur. Phys. J. C 81, 381 (2021).
  4. R. Bommasani et al., On the opportunities and risks of foundation models, Stanford CRFM report, arXiv:2108.07258.
  5. T. Golling, L. Heinrich, M. Kagan, S. Klein, M. Leigh, M. Osadchy, and J. A. Raine, Masked particle modeling on sets: Towards self-supervised high energy physics foundation models, Mach. Learn.: Sci. Technol. 5, 035074 (2024).
  6. J. Birk, A. Hallin, and G. Kasieczka, OmniJet-α: The first cross-task foundation model for particle physics, Mach. Learn.: Sci. Technol. 5, 035031 (2024).
  7. M. Leigh, S. Klein, F. Charton, T. Golling, L. Heinrich, M. Kagan, I. Ochoa, and M. Osadchy, Is tokenization needed for masked particle modelling? Mach. Learn. Sci. Tech. 6, 025075 (2025).
  8. P. Harris, J. Krupa, M. Kagan, B. Maier, and N. Woodward, Resimulation-based self-supervised learning for pretraining physics foundation models, Phys. Rev. D 111, 032010 (2025).
  9. T. Kishimoto, M. Morinaga, M. Saito, and J. Tanaka, Pre-training strategy using real particle collision data for event classification in collider physics, in NeurIPS 2023 Workshop: Machine Learning and the Physical Sciences Workshop, arXiv:2312.06909.
  10. V. Mikuni and B. Nachman, Solving key challenges in collider physics with foundation models, Phys. Rev. D 111, L051504 (2025).
  11. J. Ho, R. Roberts, S. Han, and H. Wang, Pretrained event classification model for high energy physics analysis, J. Instrum. 21, P08006 (2026).
  12. T.-H. Hsu, B.-H. Zhou, Q. Liu, Y. Xu, S. Li, G. W.-S. Hou, B. Nachman, S.-C. Hsu, V. Mikuni, Y.-T. Chou, and Y. Zhang, EveNet: A foundation model for particle collision data analysis, arXiv:2601.17126.
  13. O. Amram, L. Anzalone, J. Birk, D. A. Faroughy, A. Hallin, G. Kasieczka, M. Krämer, I. Pang, H. Reyes-Gonzalez, and D. Shih, Aspen open jets: Unlocking LHC data for foundation models in particle physics, Mach. Learn.: Sci. Technol. 6, 030601 (2025).
  14. M. Vigl, N. Hartman, and L. Heinrich, Finetuning foundation models for joint analysis optimization in high energy physics, Mach. Learn.: Sci. Technol. 5, 025075 (2024).
  15. C. Li et al., Accelerating resonance searches via signature-oriented pre-training, arXiv:2405.12972.
  16. Z. Zhao, F. Mokhtar, R. Kansal, H. Li, and J. Duarte, Large-scale pretraining and finetuning for efficient jet classification in particle physics, in Proceedings of 22nd International Workshop on Advanced Computing and Analysis Techniques inPhysics Research, arXiv:2408.09343.
  17. V. Mikuni and B. Nachman, Method to simultaneously facilitate all jet physics tasks, Phys. Rev. D 111, 054015 (2025).
  18. W. Bhimji, C. Harris, V. Mikuni, and B. Nachman, Foundation model framework for all tasks involving jet physics, Phys. Rev. D 113, 032020 (2026).
  19. A. Hallin, Foundation models for high-energy physics, in Proceedings of 2nd European AI for Fundamental Physics Conference, arXiv:2509.21434.
  20. Particle Data Group Collaboration, Machine learning, in Review of particle physics, Int. J. Mod. Phys. A 41, 2630011 (2026).
  21. J. Pata, E. Wulff, F. Mokhtar, D. Southwick, M. Zhang, M. Girone, and J. Duarte, Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors, Commun. Phys. 7, 124 (2024).
  22. F. Mokhtar, J. Pata, D. Garcia, E. Wulff, M. Zhang, M. Kagan, and J. Duarte, Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders, Phys. Rev. D 111, 092015 (2025).
  23. D. Garcia, G. Krzmanc, P. Zehetner, J. Kieseler, and M. Selvaggi, Towards particle flow event reconstruction at the Future Circular Collider with GNNs, in NeurIPS 2023 Workshop: New Frontiers in Graph Learning (2023).
  24. P. Wahlen and T. Suehara, Particle-flow reconstruction with Transformer, EPJ Web Conf. 315, 03010 (2024).
  25. D. Garcia, L. Herrmann, G. Krzmanc, and M. Selvaggi, End-to-end event reconstruction for precision physics at future colliders, arXiv:2603.04084.
  26. A. Hayrapetyan et al. (CMS Collaboration), Full event interpretation with machine-learning-based particle-flow reconstruction in the CMS detector, Eur. Phys. J. C 86, 1025 (2026).
  27. L. van der Maaten and G. Hinton, Visualizing data using t-SNE, J. Mach. Learn. Res. 9, 2579 (2008).
  28. N. Bacchetta et al., CLD—A detector concept for the FCC-ee, arXiv:1911.12230.
  29. A. Abada et al. (FCC Collaboration), FCC-ee: The lepton collider: Future Circular Collider conceptual design report volume 2, Eur. Phys. J. Spec. Top. 228, 261 (2019).
  30. C. Bierlich, S. Chakraborty, N. Desai, L. Gellersen, I. Helenius, P. Ilten, L. Lönnblad, S. Mrenna, S. Prestel, C. T. Preuss, T. Sjöstrand, P. Skands, M. Utheim, and R. Verheyen, A comprehensive guide to the physics and usage of pythia 8.3, SciPost Phys. Codebases 8 (2022).
  31. S. Agostinelli et al. (Geant4 Collaboration), geant4—A simulation toolkit, Nucl. Instrum. Methods Phys. Res. Sect. A 506, 250 (2003).
  32. J. Allison geant4 developments and applications, IEEE Trans. Nucl. Sci. 53, 270 (2006).
  33. J. Allison et al. (GEANT4 Collaboration), Recent developments in geant4, Nucl. Instrum. Methods Phys. Res. Sect. A 835, 186 (2016).
  34. F. Pedregosa et al., Scikit-learn: Machine learning in Python, J. Mach. Learn. Res. 12, 2825 (2011).
  35. G. Alain and Y. Bengio, Understanding intermediate layers using linear classifier probes, arXiv:1610.01644.
  36. M. Cacciari, G. P. Salam, and G. Soyez, The anti-kT jet clustering algorithm, J. High Energy Phys. 04 (2008) 063.
  37. M. Cacciari, G. P. Salam, and G. Soyez, The catchment area of jets, J. High Energy Phys. 04 (2008) 005 .
  38. F. Bedeschi, L. Gouskos, and M. Selvaggi, Jet flavour tagging for future colliders with fast simulation, Eur. Phys. J. C 82, 646 (2022).
  39. S. Aumiller, D. Garcia, and M. Selvaggi, “Jet flavor tagging performance at FCC-ee”, 2024, https://doi.org/10.17181/e537w-n6886.
  40. N. Bacchetta, P. Collins, and P. Riedler, Tracking and vertex detectors at FCC-ee, Eur. Phys. J. Plus 137, 231 (2022).
  41. H. Qu and L. Gouskos, ParticleNet: Jet tagging via particle clouds, Phys. Rev. D 101, 056019 (2020).
  42. L. Smarr et al., The Pacific research platform: Making high-speed networking a reality for the scientist, in Practice and Experience on Advanced Research Computing (PEARC18) (ACM Press, New York, 2018).
  43. I. Loshchilov and F. Hutter, Decoupled weight decay regularization, arXiv:1711.05101.
  44. CMS Collaboration, A unified approach for jet tagging in Run 3 at s=13.6 TeV in CMS, CMS Detector Performance Summary CMS-DP-2024-066, 2024.
  45. V. Chekhovsky et al. (CMS Collaboration), High-precision measurement of the W boson mass with the CMS experiment, Nature (London) 652, 321 (2026).
  46. A. Hayrapetyan et al. (CMS Collaboration), Improving missing transverse momentum estimation with a deep neural network, Phys. Rev. D 113, 072010 (2026).
  47. A. Paszke et al., PyTorch: An imperative style, high-performance deep learning library, in Advances in Neural Information Processing Systems, edited by H. Wallach, H. Larochelle, A. Beygelzimer, F. d’Alché-Buc, E. Fox, and R. Garnett (Curran Associates, Red Hook, NY, 2019), Vol. 32, p. 8024.
  48. D. Bertolini, P. Harris, M. Low, and N. Tran, Pileup per particle identification, J. High Energy Phys. 10 (2014) 059.
  49. F. Mokhtar, J. Pata, M. Kagan, and J. Duarte, MLPF as a foundation model—Downstream evaluation dataset (FCC-ee CLD, e+e−→tt¯, s=365 GeV) [Dataset], Zenodo (2026), https://doi.org/10.5281/zenodo.20663506.

Outline

Information

PRX Intelligence 1, 023012– Published 4 November, 2026
  • Received 15 June 2026
  • Accepted 17 September 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