pith. sign in

arxiv: 2605.15549 · v1 · pith:NJD4WUUHnew · submitted 2026-05-15 · 💻 cs.LG · cs.AI· cs.CE

CTF4Nuclear: Common Task Framework for Nuclear Fission and Fusion Models

Stefano Riva , Carolina Introini , Antonio Cammi , Dean Price , Alexey Yermakov , Yue Zhao , Philippe M. Wyder , Judah Goldfeder
show 8 more authors
Jan Williams Amy Sara Rude Matteo Tomasetto Joe Germany Joseph Bakarji Georg Maierhofer Miles Cranmer J. Nathan Kutz
This is my paper

Pith reviewed 2026-05-20 20:25 UTC · model grok-4.3

classification 💻 cs.LG cs.AIcs.CE
keywords machine learningnuclear engineeringcommon task frameworkfission modelsfusion modelssparse measurementssystem monitoringsurrogate models
0
0 comments X

The pith

A Common Task Framework standardizes machine learning comparisons for nuclear fission and fusion systems through curated datasets, twelve metrics, and a sparse-measurement monitoring task.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces a Common Task Framework to enable fair comparisons of machine learning methods applied to nuclear engineering problems. It assembles datasets from nuclear and nuclear-adjacent systems and defines evaluation through twelve established metrics plus a dedicated task that monitors system state from sparse measurements alone. This structure is intended to replace inconsistent ad hoc testing with reproducible benchmarks performed on hidden test sets. In a safety-critical field where high-fidelity simulations are too slow for real time use, such standardization would clarify which data-driven approaches can serve as reliable surrogates for complex reactor dynamics.

Core claim

We introduce a Common Task Framework for machine learning in nuclear engineering that draws on curated datasets from different nuclear and nuclear-adjacent systems. Performance is measured with twelve established metrics together with a new paradigm for system monitoring that uses sparse measurements only. Benchmarking of standard baselines on these resources reveals current method limitations, and the framework is offered to support standardized evaluations on hidden test sets that raise rigour and reproducibility for scientific machine learning in the nuclear industry.

What carries the argument

The Common Task Framework, which supplies standardized nuclear datasets, twelve performance metrics, and a sparse-measurement monitoring paradigm to produce comparable results across machine learning methods.

If this is right

  • Ad hoc comparisons among machine learning methods will give way to standardized evaluations performed on hidden test sets.
  • Rigour and reproducibility will increase for scientific machine learning applied to the nuclear industry.
  • A clearer picture will emerge of the advantages and limitations of different machine learning approaches in safety-critical nuclear settings.
  • Limitations of current baseline methods for nuclear applications will be identified more consistently through repeated use of the same benchmarks.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Widespread use of the framework could speed development of surrogate models that run fast enough for real-time nuclear system monitoring and control.
  • The sparse-measurement focus may encourage new machine learning techniques that handle incomplete data across other multi-physics domains.
  • Cross-application of methods validated on this framework to fission-only or fusion-only test cases could test how well performance generalizes within nuclear technologies.
  • Hybrid approaches that combine the framework's data-driven evaluations with existing high-fidelity physics simulators may become easier to design and validate.

Load-bearing premise

The curated nuclear datasets together with the twelve chosen metrics and the sparse-monitoring task are sufficient to support fair and meaningful comparisons of machine learning methods in safety-critical nuclear settings.

What would settle it

A machine learning method that scores well on the twelve metrics and the sparse-monitoring task yet produces inaccurate predictions when tested on independent nuclear reactor data or simulations outside the curated collection would show that the framework does not reliably identify suitable methods.

Figures

Figures reproduced from arXiv: 2605.15549 by Alexey Yermakov, Amy Sara Rude, Antonio Cammi, Carolina Introini, Dean Price, Georg Maierhofer, Jan Williams, J. Nathan Kutz, Joe Germany, Joseph Bakarji, Judah Goldfeder, Matteo Tomasetto, Miles Cranmer, Philippe M. Wyder, Stefano Riva, Yue Zhao.

Figure 1
Figure 1. Figure 1: Scheme of the Common Task Framework (CTF) for Nuclear Fission and Fusion Models, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The CTF scores the performance of methods on nuclear reactors data. (a) Data is collected [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The CTF4Nuclear introduces a new task focused on the reconstruction of the state of [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Bar plots of each task from E1 to E12, with the top 3 methods for each task. The blue [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
read the original abstract

The demand for clean energy is ever increasing, with new nuclear technologies presenting a complementary solution to renewable energies. However, designing and operating these systems is exceptionally difficult, given the complexity of the physical phenomena that interact to form the system dynamics. While high-fidelity simulations help to understand the non-linear, multi-physics interactions within a reactor, they are computationally expensive and rarely suitable for real-time applications. Furthermore, model-based approaches are inherently sensitive to simplifying assumptions required to derive their governing equations and parameters, leading to inevitable discrepancies with real-world measurements. In contrast, Machine Learning (ML) methods have the potential to generate reliable surrogate models which may be able to quickly predict the system's behaviour. However, the number of data-driven methods that can potentially be used for this task is large and diverse. In a safety-critical setting such as nuclear engineering, a fair comparison of different ML methods, and a clear understanding of their advantages and limitations, is of paramount importance. To address this, we introduce a Common Task Framework (CTF) for ML in nuclear engineering, building upon previous efforts in dynamical systems and seismology. This CTF considers a curated set of datasets from different nuclear and nuclear-adjacent systems. The CTF evaluates the performance of a method on 12 established metrics, alongside a new paradigm focused on system monitoring from sparse measurements only. We illustrate the framework by benchmarking standard ML baselines against these datasets, revealing current method limitations. Our vision is to replace ad hoc comparisons with standardized evaluations on hidden test sets, raising the bar for rigour and reproducibility in scientific ML for the nuclear industry.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 3 minor

Summary. The manuscript introduces CTF4Nuclear, a Common Task Framework for machine learning in nuclear fission and fusion applications. It supplies curated datasets drawn from nuclear and nuclear-adjacent systems, defines evaluation on 12 established metrics, and adds a new paradigm for system monitoring from sparse measurements only. Standard ML baselines are benchmarked on these datasets to illustrate current limitations, with the goal of replacing ad-hoc comparisons by standardized evaluations on hidden test sets to improve rigor and reproducibility.

Significance. If the framework is adopted and the datasets released, the work could raise the standard for reproducible ML benchmarking in safety-critical nuclear engineering. The combination of multiple metrics with a sparse-monitoring task directly targets the gap between expensive high-fidelity simulations and real-time surrogate needs, and the explicit focus on hidden test sets addresses a known weakness in current nuclear-ML literature.

major comments (1)
  1. [§4.2, Table 3] §4.2 and Table 3: the claim that the 12 metrics 'reveal current method limitations' rests on the reported baseline scores, yet no statistical significance tests, confidence intervals, or cross-validation details are provided for the sparse-monitoring task; without these the quantitative evidence for the central claim of method inadequacy is incomplete.
minor comments (3)
  1. [§3.1] §3.1: the precise criteria used to select the 'nuclear-adjacent' datasets are stated only at a high level; adding a short table or paragraph listing the physical quantities and operating regimes would improve reproducibility.
  2. [Figure 2] Figure 2: axis labels and units are missing on the sparse-measurement reconstruction plots, making it difficult to judge the practical scale of the reported errors.
  3. The manuscript does not indicate whether the curated datasets and hidden test splits will be released under an open license; this detail is essential for the framework's intended community impact.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive comments and for recommending minor revision. We address the major comment below.

read point-by-point responses

  1. Referee: [§4.2, Table 3] §4.2 and Table 3: the claim that the 12 metrics 'reveal current method limitations' rests on the reported baseline scores, yet no statistical significance tests, confidence intervals, or cross-validation details are provided for the sparse-monitoring task; without these the quantitative evidence for the central claim of method inadequacy is incomplete.

    Authors: We agree that the presentation of the baseline results would be strengthened by additional statistical details. The baselines in §4.2 and Table 3 are provided as illustrative examples to demonstrate how the CTF can be used and to highlight potential shortcomings of standard methods on these tasks, rather than as a definitive statistical ranking. In the revised version we will expand §4.2 to describe the cross-validation procedure employed for the sparse-monitoring task and will augment Table 3 with bootstrap confidence intervals for each reported metric. We will also clarify in the text that these results serve to motivate the framework and that the CTF is intended to support more rigorous future evaluations on the hidden test sets. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected in CTF proposal

full rationale

The manuscript proposes a Common Task Framework consisting of curated nuclear datasets, 12 metrics, and a sparse-monitoring evaluation paradigm to standardize ML method comparisons. This is an infrastructure and benchmarking contribution rather than a derivation of new equations or predictions; the central claim is that the released hidden test sets and metrics will enable fair external evaluation. No load-bearing step reduces a claimed result to a fitted parameter, self-definition, or self-citation chain within the paper itself. The argument is self-contained against external benchmarks and adoption, with no internal reduction of the framework's value to quantities defined inside the manuscript.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The paper is a framework proposal rather than a derivation; it relies on domain assumptions about dataset representativeness and metric suitability but introduces no fitted numerical parameters or new physical entities.

axioms (1)
  • domain assumption Curated nuclear and nuclear-adjacent datasets are representative for evaluating ML surrogate models in safety-critical settings
    Invoked when defining the CTF datasets and evaluation tasks
invented entities (1)
  • CTF4Nuclear framework no independent evidence
    purpose: To provide standardized, reproducible evaluation of ML methods for nuclear systems
    Newly defined in this work as the central contribution

pith-pipeline@v0.9.0 · 5884 in / 1347 out tokens · 59616 ms · 2026-05-20T20:25:42.167732+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

86 extracted references · 86 canonical work pages · 2 internal anchors

  1. [1]

    History, Development and Future of TRIGA Research Reactors

    International Atomic Energy Agency. History, Development and Future of TRIGA Research Reactors. Technical Report 482, IAEA, 2016

    work page 2016

  2. [2]

    Chronos-2: From Univariate to Universal Forecasting

    Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sun- dar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, et al. Chronos-2: From univariate to universal forecasting.arXiv preprint arXiv:2510.15821, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  3. [3]

    Chronos: Learning the Language of Time Series

    Abdul Fatir Ansari, Lorenzo Stella, Caner Turkmen, Xiyuan Zhang, Pedro Mercado, Huibin Shen, Olek- sandr Shchur, Syama Sundar Rangapuram, Sebastian Pineda Arango, Shubham Kapoor, et al. Chronos: Learning the language of time series.arXiv preprint arXiv:2403.07815, 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  4. [4]

    Argaud, B Bouriquet, F de Caso, H Gong, Y Maday, and O Mula

    J.-P. Argaud, B Bouriquet, F de Caso, H Gong, Y Maday, and O Mula. Sensor placement in nuclear reactors based on the generalized empirical interpolation method.Journal of Computational Physics, 363:354–370, 2018

    work page 2018

  5. [5]

    Variable projection methods for an optimized dynamic mode decompo- sition.SIAM Journal on Applied Dynamical Systems, 17(1):380–416, 2018

    Travis Askham and J Nathan Kutz. Variable projection methods for an optimized dynamic mode decompo- sition.SIAM Journal on Applied Dynamical Systems, 17(1):380–416, 2018

    work page 2018

  6. [6]

    Ricotti, and Hervé Rouch

    Manuele Aufiero, Antonio Cammi, Olivier Geoffroy, Mario Losa, Lelio Luzzi, Marco E. Ricotti, and Hervé Rouch. Development of an OpenFOAM model for the Molten Salt Fast Reactor transient analysis. Chemical Engineering Science, 111:390–401, May 2014

    work page 2014

  7. [7]

    M. G. Bellemare, Y . Naddaf, J. Veness, and M. Bowling. The arcade learning environment: An evaluation platform for general agents.Journal of Artificial Intelligence Research, 47:253–279, June 2013

    work page 2013

  8. [8]

    Openai gym, 2016

    Greg Brockman, Vicki Cheung, Ludwig Pettersson, Jonas Schneider, John Schulman, Jie Tang, and Wojciech Zaremba. Openai gym, 2016

    work page 2016

  9. [9]

    Optimization of the pre-conceptual design of the MSFR, 2013

    Mariya Brovchenko, Elsa Merle Lucotte, Hervé Rouch, Fabio Alcaro, M Allibert, M Aufiero, Antonio Cammi, S Dulla, O Feynberg, L Frima, et al. Optimization of the pre-conceptual design of the MSFR, 2013

    work page 2013

  10. [10]

    Artificial intelligence for accelerating nuclear applications, science, and technology

    David Brown. Artificial intelligence for accelerating nuclear applications, science, and technology. Technical report, Brookhaven National Laboratory (BNL), Upton, NY (United States), 07 2022

    work page 2022

  11. [11]

    Brunton, Marko Budiši´c, Eurika Kaiser, and J

    Steven L. Brunton, Marko Budiši´c, Eurika Kaiser, and J. Nathan Kutz. Modern Koopman Theory for Dynamical Systems.SIAM Review, 64(2):229–340, 2022

    work page 2022

  12. [12]

    Brunton, Joshua L

    Steven L. Brunton, Joshua L. Proctor, and J. Nathan Kutz. Discovering governing equations from data by sparse identification of nonlinear dynamical systems.Proceedings of the National Academy of Sciences, 113(15):3932–3937, 2016

    work page 2016

  13. [13]

    Codbench: a critical evaluation of data-driven models for continuous dynamical systems.Digital Discovery, 3(6):1172– 1181, 2024

    Priyanshu Burark, Karn Tiwari, Meer Mehran Rashid, Prathosh AP, and NM Anoop Krishnan. Codbench: a critical evaluation of data-driven models for continuous dynamical systems.Digital Discovery, 3(6):1172– 1181, 2024

    work page 2024

  14. [15]

    Antonio Cammi, Stefano Riva, Carolina Introini, Lorenzo Loi, and Enrico Padovani. Data-driven model order reduction for sensor positioning and indirect reconstruction with noisy data: Application to a Circulating Fuel Reactor.Nuclear Engineering and Design, 421:113105, May 2024

    work page 2024

  15. [16]

    Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde de Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, Alex Ray, Raul Puri, Gretchen Krueger, Michael Petrov, Heidy Khlaaf, Girish Sastry, Pamela Mishkin, Brooke Chan, Scott Gray, Nick Ryder, Mikhail Pavlov, Alethea Power, Lukasz Kaiser, Mohammad Bavarian...

    work page 2021

  16. [17]

    Training verifiers to solve math word problems, 2021

    Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. Training verifiers to solve math word problems, 2021

    work page 2021

  17. [18]

    Coelho, M

    C. Coelho, M. Fernanda P. Costa, and Luis L. Ferrás. Enhancing continuous time series modelling with a latent ode-lstm approach.Applied Mathematics and Computation, 475:128727, 2024

    work page 2024

  18. [19]

    6 - Neutronic/thermal-hydraulic coupling

    Christophe Demazière. 6 - Neutronic/thermal-hydraulic coupling. In Christophe Demazière, editor, Modelling of Nuclear Reactor Multi-physics, pages 311–336. Academic Press, 2020

    work page 2020

  19. [20]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In2009 IEEE Conference on Computer Vision and Pattern Recognition, pages 248–255. IEEE, 2009

    work page 2009

  20. [21]

    50 years of data science.Journal of Computational and Graphical Statistics, 26(4):745–766, 2017

    David Donoho. 50 years of data science.Journal of Computational and Graphical Statistics, 26(4):745–766, 2017

    work page 2017

  21. [22]

    Duderstadt and Louis J

    James J. Duderstadt and Louis J. Hamilton.Nuclear Reactor Analysis. John Wiley & Sons, New York, 1976

    work page 1976

  22. [23]

    Shallow recurrent decoder for reduced order modeling of e × b plasma dynamics.Machine Learning: Science and Technology, 6(2):025024, apr 2025

    Farbod Faraji, Maryam Reza, and J Nathan Kutz. Shallow recurrent decoder for reduced order modeling of e × b plasma dynamics.Machine Learning: Science and Technology, 6(2):025024, apr 2025

    work page 2025

  23. [24]

    Nathan Kutz, Bingni W

    Urban Fasel, J. Nathan Kutz, Bingni W. Brunton, and Steven L. Brunton. Ensemble-sindy: Robust sparse model discovery in the low-data, high-noise limit, with active learning and control.Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 478(2260):20210904, 2022

    work page 2022

  24. [25]

    Impact assessment of radiative heat transport in arc-class reactor flibe liquid immersion blanket.Nuclear Science and Engineering, 198(4):898–913, 2024

    Gabriele Ferrero, Raffaella Testoni, and Massimo Zucchetti. Impact assessment of radiative heat transport in arc-class reactor flibe liquid immersion blanket.Nuclear Science and Engineering, 198(4):898–913, 2024

    work page 2024

  25. [26]

    Physics-informed neural network for solving neutron transport equation with differential order transformation and fourier feature.Journal of Computational Physics, 546:114519, 2026

    Xiaoxu Geng, Dandan Chen, Jiayi Yu, Chong Wang, Changjun Hu, and Zhaoshun Wang. Physics-informed neural network for solving neutron transport equation with differential order transformation and fourier feature.Journal of Computational Physics, 546:114519, 2026

    work page 2026

  26. [27]

    Péter German, Mauricio Tano, Carlo Fiorina, and Jean C. Ragusa. Gen-rom—an openfoam®-based multiphysics reduced-order modeling framework for the analysis of molten salt reactors.Progress in Nuclear Energy, 146:104148, 2022

    work page 2022

  27. [28]

    Ai must embrace specialization via superhuman adaptable intelligence.arXiv preprint arXiv:2602.23643, 2026

    Judah Goldfeder, Philippe Wyder, Yann LeCun, and Ravid Shwartz Ziv. Ai must embrace specialization via superhuman adaptable intelligence.arXiv preprint arXiv:2602.23643, 2026

    work page arXiv 2026

  28. [29]

    Generalized Empirical Interpolation Method With H1 Regulariza- tion: Application to Nuclear Reactor Physics.Frontiers in Energy Research, 9, 2022

    Helin Gong, Zhang Chen, and Qing Li. Generalized Empirical Interpolation Method With H1 Regulariza- tion: Application to Nuclear Reactor Physics.Frontiers in Energy Research, 9, 2022

    work page 2022

  29. [30]

    Parameter identification and state estimation for nuclear reactor operation digital twin.Annals of Nuclear Energy, 180:109497, January 2023

    Helin Gong, Tao Zhu, Zhang Chen, Yaping Wan, and Qing Li. Parameter identification and state estimation for nuclear reactor operation digital twin.Annals of Nuclear Energy, 180:109497, January 2023

    work page 2023

  30. [31]

    Plasma surrogate modelling using Fourier neural operators.arXiv preprint arXiv:2311.05967, 2024

    Vignesh Gopakumar, Stanislas Pamela, Lorenzo Zanisi, Zongyi Li, Ander Gray, Daniel Brennand, Nitesh Bhatia, Gregory Stathopoulos, Matt Kusner, Marc Peter Deisenroth, Anima Anandkumar, JOREK Team, and MAST Team. Plasma surrogate modelling using Fourier neural operators.arXiv preprint arXiv:2311.05967, 2024

    work page arXiv 2024

  31. [32]

    Grieves.Virtually Intelligent Product Systems: Digital and Physical Twins, pages 175–200

    Michael W. Grieves.Virtually Intelligent Product Systems: Digital and Physical Twins, pages 175–200. Progress in Astronautics and Aeronauting, 2019

    work page 2019

  32. [33]

    Large language models are zero-shot time series forecasters.Advances in Neural Information Processing Systems, 36:19622–19635, 2023

    Nate Gruver, Marc Finzi, Shikai Qiu, and Andrew G Wilson. Large language models are zero-shot time series forecasters.Advances in Neural Information Processing Systems, 36:19622–19635, 2023

    work page 2023

  33. [34]

    Deepseek-r1 incentivizes reasoning in llms through reinforcement learning.Nature, 645(8081):633–638, 09 2025

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, et al. Deepseek-r1 incentivizes reasoning in llms through reinforcement learning.Nature, 645(8081):633–638, 09 2025

    work page 2025

  34. [35]

    Measuring mathematical problem solving with the math dataset, 2021

    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, 2021

    work page 2021

  35. [36]

    Long short-term memory.Neural computation, 9(8):1735–1780, 1997

    Sepp Hochreiter and Jürgen Schmidhuber. Long short-term memory.Neural computation, 9(8):1735–1780, 1997

    work page 1997

  36. [37]

    From tables to time: How tabpfn-v2 outperforms specialized time series forecasting models, 2025

    Shi Bin Hoo, Samuel Müller, David Salinas, and Frank Hutter. From tables to time: How tabpfn-v2 outperforms specialized time series forecasting models, 2025. 11

    work page 2025

  37. [38]

    Virtual sensing-enabled digital twin framework for real-time monitoring of nuclear systems leveraging deep neural operators.npj Materials Degradation, 9(1):21, 2025

    Raisa Hossain, Farid Ahmed, Kazuma Kobayashi, Seid Koric, Diab Abueidda, and Syed Bahauddin Alam. Virtual sensing-enabled digital twin framework for real-time monitoring of nuclear systems leveraging deep neural operators.npj Materials Degradation, 9(1):21, 2025

    work page 2025

  38. [39]

    A review of the application of artificial intelligence to nuclear reactors: Where we are and what’s next.Heliyon, 9(3):e13883, 2023

    Qingyu Huang, Shinian Peng, Jian Deng, Hui Zeng, Zhuo Zhang, Yu Liu, and Peng Yuan. A review of the application of artificial intelligence to nuclear reactors: Where we are and what’s next.Heliyon, 9(3):e13883, 2023

    work page 2023

  39. [40]

    Advances in small modular reactor technology developments

    IAEA. Advances in small modular reactor technology developments. Technical report, International Atomic Energy Agency, Vienna, 2023

    work page 2023

  40. [41]

    the ’echo state’ approach to analyzing and training recurrent neural networks

    Herbet Jaeger. "the ’echo state’ approach to analyzing and training recurrent neural networks". Technical report, German National Research Center for Information Technology, Technical Report GMD 148, 2001

    work page 2001

  41. [42]

    John Jumper, Richard Evans, Alexander Pritzel, Tim Green, Michael Figurnov, Olaf Ronneberger, Kathryn Tunyasuvunakool, Russ Bates, Augustin Žídek, Anna Potapenko, Alex Bridgland, Clemens Meyer, Simon A. A. Kohl, Andrew J. Ballard, Andrew Cowie, Bernardino Romera-Paredes, Stanislav Nikolov, Rishub Jain, Jonas Adler, Trevor Back, Stig Petersen, David Reiman...

    work page 2021

  42. [43]

    Kapteyn, David J

    Michael G. Kapteyn, David J. Knezevic, and Karen Willcox. Toward predictive digital twins via component- based reduced-order models and interpretable machine learning. InAIAA Scitech 2020 Forum, 2020. _eprint: https://arc.aiaa.org/doi/pdf/10.2514/6.2020-0418

    work page doi:10.2514/6.2020-0418 2020

  43. [44]

    Deep neural operator-driven real-time inference to enable digital twin solutions for nuclear energy systems.Scientific Reports, 14(1):2101, 2024

    Kazuma Kobayashi and Syed Bahauddin Alam. Deep neural operator-driven real-time inference to enable digital twin solutions for nuclear energy systems.Scientific Reports, 14(1):2101, 2024

    work page 2024

  44. [45]

    Imagenet classification with deep convolutional neural networks

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. InAdvances in Neural Information Processing Systems, pages 1097–1105, 2012

    work page 2012

  45. [46]

    Panda: A pretrained forecast model for chaotic dynamics

    Jeffrey Lai, Anthony Bao, and William Gilpin. Panda: A pretrained forecast model for chaotic dynamics. arXiv preprint arXiv:2505.13755, 2025

    work page arXiv 2025

  46. [47]

    Model Order Reduction in Fluid Dynamics: Challenges and Perspectives

    Toni Lassila, Andrea Manzoni, Alfio Quarteroni, and Gianluigi Rozza. Model Order Reduction in Fluid Dynamics: Challenges and Perspectives. InReduced Order Methods for Modeling and Computational Reduction, pages 235–273. Springer International Publishing, Cham, 2014

    work page 2014

  47. [48]

    Fourier neural operator for partial differential equations

    Zongyi Li, Nikola Kovachki, Kamyar Azizzadenesheli, Burigede Liu, Kaushik Bhattacharya, Andrew Stuart, and Anima Anandkumar. Fourier neural operator for partial differential equations. InInternational Conference on Learning Representations (ICLR), 2021

    work page 2021

  48. [49]

    Fourier neural operator for parametric partial differential equations

    Zongyi Li, Nikola Borislavov Kovachki, Kamyar Azizzadenesheli, Burigede liu, Kaushik Bhattacharya, Andrew Stuart, and Anima Anandkumar. Fourier neural operator for parametric partial differential equations. InInternational Conference on Learning Representations, 2021

    work page 2021

  49. [50]

    Moirai 2.0: When less is more for time series forecasting.arXiv preprint arXiv:2511.11698, 2025

    Chenghao Liu, Taha Aksu, Juncheng Liu, Xu Liu, Hanshu Yan, Quang Pham, Silvio Savarese, Doyen Sahoo, Caiming Xiong, and Junnan Li. Moirai 2.0: When less is more for time series forecasting.arXiv preprint arXiv:2511.11698, 2025

    work page arXiv 2025

  50. [51]

    Moirai-moe: Empowering time series foundation models with sparse mixture of experts.arXiv preprint arXiv:2410.10469, 2024

    Xu Liu, Juncheng Liu, Gerald Woo, Taha Aksu, Yuxuan Liang, Roger Zimmermann, Chenghao Liu, Silvio Savarese, Caiming Xiong, and Doyen Sahoo. Moirai-moe: Empowering time series foundation models with sparse mixture of experts.arXiv preprint arXiv:2410.10469, 2024

    work page arXiv 2024

  51. [52]

    Sundial: A family of highly capable time series foundation models, 2025

    Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, and Mingsheng Long. Sundial: A family of highly capable time series foundation models, 2025

    work page 2025

  52. [53]

    Hou, and Max Tegmark

    Ziming Liu, Yixuan Wang, Sachin Vaidya, Fabian Ruehle, James Halverson, Marin Soljaˇci´c, Thomas Y . Hou, and Max Tegmark. Kan: Kolmogorov-arnold networks, 2025

    work page 2025

  53. [54]

    A novel openfoam library for magneto-hydrodynamics studies in the nuclear fusion field

    Matteo Lo Verso, Carolina Introini, Eric Cervi, Francesca Giacobbo, Luciana Barucca, Marco Caramello, Matteo Prinzio, Laura Savoldi, and Antonio Cammi. A novel openfoam library for magneto-hydrodynamics studies in the nuclear fusion field. InProceedings of NUTHOS-14 Conference, 08 2024

    work page 2024

  54. [55]

    Learning nonlinear operators via deeponet based on the universal approximation theorem of operators.Nature Machine Intelligence, 3(3):218–229, 2021

    Lu Lu, Pengzhan Jin, Guofei Pang, Zhongqiang Zhang, and George Em Karniadakis. Learning nonlinear operators via deeponet based on the universal approximation theorem of operators.Nature Machine Intelligence, 3(3):218–229, 2021. 12

    work page 2021

  55. [56]

    Learning nonlinear operators via deeponet based on the universal approximation theorem of operators.Nature Machine Intelligence, 3(3):218–229, 2021

    Lu Lu, Pengzhan Jin, Guofei Pang, Zhongqiang Zhang, and George Em Karniadakis. Learning nonlinear operators via deeponet based on the universal approximation theorem of operators.Nature Machine Intelligence, 3(3):218–229, 2021

    work page 2021

  56. [57]

    Yantao Luo, Carolina Introini, Mahdi Nasr Aghili, Lorenzo Loi, Andrea Salvini, Xiang Wang, Stefano Riva, and Antonio Cammi. Development and validation of a high-fidelity multi-physics and multi-region solver with application to the TRIGA mark II reactor.under review at Progress in Nuclear Energy, March 2026

    work page 2026

  57. [58]

    On the computational power of circuits of spiking neurons.Journal of Computer and System Sciences, 69(4):593–616, December 2004

    Wolfgang Maass and Henry Markram. On the computational power of circuits of spiking neurons.Journal of Computer and System Sciences, 69(4):593–616, December 2004

    work page 2004

  58. [59]

    Y Maday, O Mula, A T Patera, and M Yano. The Generalized Empirical Interpolation Method: Stability theory on Hilbert spaces with an application to the Stokes equation.Computer Methods in Applied Mechanics and Engineering, 287:310–334, 2015. Publisher: Elsevier B.V

    work page 2015

  59. [60]

    Yvon Maday, Anthony Patera, James Penn, and Masayuki Yano. A parameterized-background data-weak approach to variational data assimilation: formulation, analysis, and application to acoustics.International Journal for Numerical Methods in Engineering, 102, 2014

    work page 2014

  60. [61]

    Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equations.Nature Machine Intelligence, 6(10):1256–1269, Oct 2024

    Nick McGreivy and Ammar Hakim. Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equations.Nature Machine Intelligence, 6(10):1256–1269, Oct 2024

    work page 2024

  61. [62]

    Modeling and validation of the dynasty natural circulation loop with the relap5 code

    Andrea Missaglia, Gabriele Benzoni, Carolina Introini, Marco Enrico Ricotti, Antonio Cammi, and Francois Foulon. Modeling and validation of the dynasty natural circulation loop with the relap5 code. Results in Engineering, 28:107466, 2025

    work page 2025

  62. [63]

    Subhasish Mohanty and Richard B. Vilim. Physics-Infused AI/ML Based Digital-Twin Framework for Flow-Induced-Vibration Damage Prediction in a Nuclear Reactor Heat Exchanger. Technical report, Argonne National Lab. (ANL), 2021

    work page 2021

  63. [64]

    Accessed May 2026

    OECD/NEA.https://www.oecd-nea.org/jcms/pl_77779, 2024. Accessed May 2026

    work page 2024

  64. [65]

    The well: a large-scale collection of diverse physics simulations for machine learning.Advances in Neural Information Processing Systems, 37:44989–45037, 2024

    Ruben Ohana, Michael McCabe, Lucas Meyer, Rudy Morel, Fruzsina Agocs, Miguel Beneitez, Marsha Berger, Blakesly Burkhart, Stuart Dalziel, Drummond Fielding, et al. The well: a large-scale collection of diverse physics simulations for machine learning.Advances in Neural Information Processing Systems, 37:44989–45037, 2024

    work page 2024

  65. [66]

    de Silva, J

    Shaowu Pan, Eurika Kaiser, Brian M. de Silva, J. Nathan Kutz, and Steven L. Brunton. PyKoopman: A Python Package for Data-Driven Approximation of the Koopman Operator.Journal of Open Source Software, 9(94):5881, 2024

    work page 2024

  66. [67]

    Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach.Physical Review Letters, 120(2):024102, January 2018

    Jaideep Pathak, Brian Hunt, Michelle Girvan, Zhixin Lu, and Edward Ott. Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach.Physical Review Letters, 120(2):024102, January 2018

    work page 2018

  67. [68]

    An Experimental Facility to Investigate the Natural Circulation Dynamics in Presence of Distributed Heat Sources.Energy Procedia, 101:10–17, 2016

    A Pini, A Cammi, M Cauzzi, F Fanale, and L Luzzi. An Experimental Facility to Investigate the Natural Circulation Dynamics in Presence of Distributed Heat Sources.Energy Procedia, 101:10–17, 2016

    work page 2016

  68. [69]

    Tsoukalas, and Alexander Heifetz

    Konstantinos Prantikos, Stylianos Chatzidakis, Lefteri H. Tsoukalas, and Alexander Heifetz. Physics- informed neural network with transfer learning (tl-pinn) based on domain similarity measure for prediction of nuclear reactor transients.Scientific Reports, 13(1):16840, 2023

    work page 2023

  69. [70]

    Tsoukalas, and Alexander Heifetz

    Konstantinos Prantikos, Lefteri H. Tsoukalas, and Alexander Heifetz. Physics-informed neural network solution of point kinetics equations for a nuclear reactor digital twin.Energies, 15(20), 2022

    work page 2022

  70. [71]

    Prince, Joshua T

    Zachary M. Prince, Joshua T. Hanophy, Vincent M. Labouré, Yaqi Wang, Logan H. Harbour, and Namjae Choi. Neutron transport methods for multiphysics heterogeneous reactor core simulation in griffin.Annals of Nuclear Energy, 200:110365, 2024

    work page 2024

  71. [72]

    Maziar Raissi, Paris Perdikaris, and George E Karniadakis. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations.Journal of Computational physics, 378:686–707, 2019

    work page 2019

  72. [73]

    Neutron Flux Reconstruction from Out-Core Sparse Measurements using Data-Driven Reduced Order Modelling

    Stefano Riva, Sophie Deanesi, Carolina Introini, Stefano Lorenzi, and Antonio Cammi. Neutron Flux Reconstruction from Out-Core Sparse Measurements using Data-Driven Reduced Order Modelling. In Proceedings of the International Conference on Physics of Reactors, PHYSOR 2024, page 1632 – 1641, 2024. 13

    work page 2024

  73. [74]

    Nathan Kutz

    Stefano Riva, Carolina Introini, Antonio Cammi, and J. Nathan Kutz. Robust state estimation from partial out-core measurements with shallow recurrent decoder for nuclear reactors.Progress in Nuclear Energy, 189:105928, 2025

    work page 2025

  74. [75]

    De Gruyter, 2020

    Gianluigi Rozza, Martin Hess, Giovanni Stabile, Marco Tezzele, Francesco Ballarin, Carmen Gräßle, Michael Hinze, Stefan V olkwein, Francisco Chinesta, Pierre Ladeveze, Yvon Maday, Anthony Patera, and J Farhat Char.Model Order Reduction: Volume 2: Snapshot-Based Methods and Algorithms. De Gruyter, 2020

    work page 2020

  75. [76]

    Ruth, Owen R

    Mark F. Ruth, Owen R. Zinaman, Mark Antkowiak, Richard D. Boardman, Robert S. Cherry, and Morgan D. Bazilian. Nuclear-renewable hybrid energy systems: Opportunities, interconnections, and needs.Energy Conversion and Management, 78:684–694, 2014

    work page 2014

  76. [77]

    Differential equations for continuous-time deep learning.arXiv preprint arXiv:2401.03965, 2024

    Lars Ruthotto. Differential equations for continuous-time deep learning.arXiv preprint arXiv:2401.03965, 2024

    work page arXiv 2024

  77. [78]

    A general reinforcement learning algorithm that masters chess, shogi, and go through self-play.Science, 362(6419):1140–1144, 2018

    David Silver, Thomas Hubert, Julian Schrittwieser, Ioannis Antonoglou, et al. A general reinforcement learning algorithm that masters chess, shogi, and go through self-play.Science, 362(6419):1140–1144, 2018

    work page 2018

  78. [79]

    Pdebench: An extensive benchmark for scientific machine learning.Advances in Neural Information Processing Systems, 35:1596–1611, 2022

    Makoto Takamoto, Timothy Praditia, Raphael Leiteritz, Daniel MacKinlay, Francesco Alesiani, Dirk Pflüger, and Mathias Niepert. Pdebench: An extensive benchmark for scientific machine learning.Advances in Neural Information Processing Systems, 35:1596–1611, 2022

    work page 2022

  79. [80]

    Williams, Francesco Braghin, Andrea Manzoni, and J

    Matteo Tomasetto, Jan P. Williams, Francesco Braghin, Andrea Manzoni, and J. Nathan Kutz. Reduced order modeling with shallow recurrent decoder networks.Nature Communications, 16(1):10260, 2025

    work page 2025

  80. [81]

    Jiangyu Wang, Xingjie Peng, Zhang Chen, Bingyan Zhou, Yajin Zhou, and Nan Zhou. Surrogate modeling for neutron diffusion problems based on conservative physics-informed neural networks with boundary conditions enforcement.Annals of Nuclear Energy, 176:109234, 2022

    work page 2022

Showing first 80 references.