pith. machine review for the scientific record. sign in

arxiv: 2604.03511 · v1 · submitted 2026-04-03 · ✦ hep-ph

Recognition: no theorem link

Monte Carlo Event Generation with Continuous Normalizing Flows

Bernhard Schmitzer, Enrico Bothmann, Fabian Sinz, Max Knobbe, Timo Jan{\ss}en

Authors on Pith no claims yet

Pith reviewed 2026-05-13 17:48 UTC · model grok-4.3

classification ✦ hep-ph
keywords continuous normalizing flowsmonte carlo event generationphase space samplingunweightingmatrix elementsflow matchingcollider physicsmachine learning
0
0 comments X

The pith

Continuous normalizing flows achieve up to 184-fold unweighting efficiency gains in Monte Carlo event generation for high-jet collider processes.

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

The paper applies continuous normalizing flows trained with flow matching to improve phase-space sampling in Monte Carlo simulations for particle physics. For lepton pair and top quark pair production with multiple jets, these flows remap random numbers to better match the matrix element probability densities. This leads to much higher rates of accepted events after unweighting compared to standard uniform sampling. The gains are largest at high jet multiplicities where traditional methods struggle most. Combining with faster coupling layer flows gives overall wall-time speedups of about ten.

Core claim

Helicity-conditioned continuous normalizing flows trained via the flow matching method remap the random numbers used in matrix element evaluation for processes like lepton-pair and top-pair production with multiple jets. Compared to standard methods, unweighting efficiency improves by factors of up to 184 and 25 at the highest jet numbers for the two processes. Using a hybrid RegFlow approach that combines continuous flows with coupling layers yields parton-level unweighted event generation walltime gains of about a factor of ten at the highest jet numbers.

What carries the argument

Helicity-conditioned continuous normalizing flows trained with flow matching to remap random numbers to the target matrix-element density.

Load-bearing premise

The samples generated by the trained flows must match the target distribution closely enough to avoid introducing any bias in calculated physics observables.

What would settle it

A comparison showing statistically significant differences in predicted distributions of physical quantities such as transverse momentum spectra between events generated with the flows and those from standard methods.

Figures

Figures reproduced from arXiv: 2604.03511 by Bernhard Schmitzer, Enrico Bothmann, Fabian Sinz, Max Knobbe, Timo Jan{\ss}en.

Figure 1
Figure 1. Figure 1: FIG. 1. Relative improvements for the unweighting efficiencies [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
read the original abstract

We apply Continuous Normalizing Flows trained with the Flow Matching method to the problem of phase-space sampling in Monte Carlo event generation for high-energy collider physics. Focusing on lepton-pair and top quark pair production with multiple jets, the two computationally most expensive processes at the Large Hadron Collider, we train helicity-conditioned Continuous Normalizing Flows to remap the random numbers used in matrix element evaluation. Compared to standard methods, we achieve unweighting efficiency improvements by factors of up to 184 and 25 for the two processes at their respective highest jet number, at the cost of an increased evaluation time. When combining the advantages of Continuous Normalizing Flows with the fast evaluation times of Coupling Layer based Flows, using the RegFlow approach, we find parton-level unweighted event generation walltime gains of about a factor of ten at the highest jet numbers. These substantial gains highlight the promise of samplers based on machine learning for next-generation collider experiments.

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

2 major / 2 minor

Summary. The manuscript applies Continuous Normalizing Flows trained via the Flow Matching method to phase-space sampling for Monte Carlo event generation. It focuses on lepton-pair and top-quark pair production with multiple jets, training helicity-conditioned CNFs to remap random numbers, and reports unweighting efficiency gains of up to 184 and 25 for the highest jet multiplicities along with wall-time improvements of roughly a factor of ten when using the hybrid RegFlow approach that combines CNFs with coupling-layer flows.

Significance. If the central claims hold, the work offers a practical advance for addressing the computational cost of event generation for high-multiplicity processes at the LHC. The reported efficiency factors are substantial, the hybrid RegFlow strategy usefully combines the strengths of different flow architectures, and the empirical results are grounded in explicit training protocols and helicity conditioning. These elements make the approach potentially valuable for next-generation collider phenomenology.

major comments (2)
  1. [Results] The efficiency numbers (up to 184 and 25) and the RegFlow wall-time gain of ~10 rest on the trained flows producing samples whose density matches the target matrix-element distribution to high accuracy. Explicit quantitative validation—such as Kolmogorov-Smirnov tests, comparisons of differential distributions for key observables, or bias checks on acceptance rates—should be shown in the results section to confirm that no systematic bias is introduced into downstream physics quantities.
  2. [Methods] The description of how the target density is approximated during training and the precise implementation of the flow-matching loss (including any regularization or conditioning details) needs to be expanded in the methods section so that the reported efficiency factors can be reproduced independently.
minor comments (2)
  1. [Abstract] The abstract states an 'increased evaluation time' without quantifying the overhead; adding a brief numerical comparison of evaluation times would clarify the practical trade-off.
  2. [Introduction] Ensure that all acronyms (CNF, RegFlow, etc.) are defined at first use in the main text.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper applies Continuous Normalizing Flows trained with Flow Matching to remap random numbers for matrix-element phase-space sampling in two collider processes. Reported unweighting efficiency gains (up to 184 and 25) and RegFlow wall-time improvements (~10) are presented as direct empirical outcomes of training and benchmarking against standard methods, with no equations or steps that reduce these metrics to fitted parameters by construction. No self-definitional mappings, fitted inputs relabeled as predictions, load-bearing self-citations, or ansatzes smuggled via prior work appear in the argument structure. The central claims rest on the standard flow-matching objective and explicit efficiency measurements, which remain independent of the target results.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The approach rests on the standard assumption that a sufficiently expressive flow can approximate the target phase-space density after training; no additional free parameters, axioms, or invented entities are introduced beyond the usual neural-network training setup.

pith-pipeline@v0.9.0 · 5468 in / 1164 out tokens · 130326 ms · 2026-05-13T17:48:41.190727+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Open LHC Monte Carlo Event Generation

    hep-ph 2026-05 unverdicted novelty 2.0

    A review of initiatives to make LHC Monte Carlo event generations available as open data to minimize redundant simulations and resource use.

Reference graph

Works this paper leans on

62 extracted references · 62 canonical work pages · cited by 1 Pith paper · 10 internal anchors

  1. [1]

    Aadet al.(ATLAS), The ATLAS Experiment at the CERN Large Hadron Collider, JINST3(08), S08003

    G. Aadet al.(ATLAS), The ATLAS Experiment at the CERN Large Hadron Collider, JINST3(08), S08003

    work page

  2. [2]

    Chatrchyanet al.(CMS), The CMS Experiment at the CERN LHC, JINST3(08), S08004

    S. Chatrchyanet al.(CMS), The CMS Experiment at the CERN LHC, JINST3(08), S08004

    work page

  3. [3]

    Abadaet al.(FCC), FCC Physics Opportunities: Fu- ture Circular Collider Conceptual Design Report Volume 1, Eur

    A. Abadaet al.(FCC), FCC Physics Opportunities: Fu- ture Circular Collider Conceptual Design Report Volume 1, Eur. Phys. J. C79, 474 (2019)

    work page 2019

  4. [4]

    Abadaet al.(FCC), FCC-ee: The Lepton Collider: Future Circular Collider Conceptual Design Report Vol- ume 2, Eur

    A. Abadaet al.(FCC), FCC-ee: The Lepton Collider: Future Circular Collider Conceptual Design Report Vol- ume 2, Eur. Phys. J. ST228, 261 (2019)

    work page 2019

  5. [5]

    Abadaet al.(FCC), FCC-hh: The Hadron Collider: Future Circular Collider Conceptual Design Report Vol- ume 3, Eur

    A. Abadaet al.(FCC), FCC-hh: The Hadron Collider: Future Circular Collider Conceptual Design Report Vol- ume 3, Eur. Phys. J. ST228, 755 (2019)

    work page 2019

  6. [6]

    Buckleyet al., General-purpose event genera- tors for LHC physics, Phys

    A. Buckleyet al., General-purpose event genera- tors for LHC physics, Phys. Rept.504, 145 (2011), arXiv:1101.2599 [hep-ph]

    work page arXiv 2011

  7. [7]

    European Strategy Group,2020 Update of the European Strategy for Particle Physics(CERN Council, Geneva, 2020)

    work page 2020

  8. [8]

    Narainet al., The Future of US Particle Physics - The Snowmass 2021 Energy Frontier Report (2022), arXiv:2211.11084 [hep-ex]

    M. Narainet al., The Future of US Particle Physics - The Snowmass 2021 Energy Frontier Report (2022), arXiv:2211.11084 [hep-ex]

    work page arXiv 2021

  9. [9]

    Albrechtet al.(HEP Software Foundation), A Roadmap for HEP Software and Computing R&D for the 2020s, Comput

    J. Albrechtet al.(HEP Software Foundation), A Roadmap for HEP Software and Computing R&D for the 2020s, Comput. Softw. Big Sci.3, 7 (2019), arXiv:1712.06982 [physics.comp-ph]

    work page arXiv 2019

  10. [10]

    Amorosoet al.(HSF Physics Event Generator WG), Challenges in Monte Carlo Event Generator Software for High-Luminosity LHC, Comput

    S. Amorosoet al.(HSF Physics Event Generator WG), Challenges in Monte Carlo Event Generator Software for High-Luminosity LHC, Comput. Softw. Big Sci.5, 12 (2021), arXiv:2004.13687 [hep-ph]

    work page arXiv 2021

  11. [11]

    G. Aadet al.(ATLAS), Measurement of the production cross section for a Higgs boson in association with a vec- tor boson in theH→W W ∗ →ℓνℓνchannel inppcolli- sions at √s= 13 TeV with the ATLAS detector, Phys. Lett. B798, 134949 (2019), arXiv:1903.10052 [hep-ex]

    work page arXiv 2019

  12. [12]

    G. Aadet al.(ATLAS), Measurement of the associated production of a Higgs boson decaying intob-quarks with a vector boson at high transverse momentum inppcol- lisions at √s= 13 TeV with the ATLAS detector, Phys. Lett. B816, 136204 (2021), arXiv:2008.02508 [hep-ex]

    work page arXiv 2021

  13. [13]

    Aaboudet al.(ATLAS), Measurements of top-quark pair spin correlations in theeµchannel at √s= 13 TeV usingppcollisions in the ATLAS detector, Eur

    M. Aaboudet al.(ATLAS), Measurements of top-quark pair spin correlations in theeµchannel at √s= 13 TeV usingppcollisions in the ATLAS detector, Eur. Phys. J. C80, 754 (2020), arXiv:1903.07570 [hep-ex]

    work page arXiv 2020

  14. [14]

    Aadet al.(ATLAS), Measurement of thet ¯tproduc- tion cross-section in the lepton+jets channel at √s= 13 TeV with the ATLAS experiment, Phys

    G. Aadet al.(ATLAS), Measurement of thet ¯tproduc- tion cross-section in the lepton+jets channel at √s= 13 TeV with the ATLAS experiment, Phys. Lett. B810, 135797 (2020), arXiv:2006.13076 [hep-ex]

    work page arXiv 2020

  15. [15]

    G. Aadet al.(ATLAS), Modelling and computational im- provements to the simulation of single vector-boson plus jet processes for the ATLAS experiment, JHEP2022(8), 089, arXiv:2112.09588 [hep-ex]

    work page arXiv

  16. [16]

    Bothmann, A

    E. Bothmann, A. Buckley, I. A. Christidi, C. G¨ utschow, S. H¨ oche, M. Knobbe, T. Martin, and M. Sch¨ onherr, Accelerating LHC event generation with simplified pi- lot runs and fast PDFs, Eur. Phys. J. C82, 1128 (2022), arXiv:2209.00843 [hep-ph]

    work page arXiv 2022

  17. [17]

    Kleiss and R

    R. Kleiss and R. Pittau, Weight optimization in multi- channel Monte Carlo, Comput. Phys. Commun.83, 141 (1994), arXiv:hep-ph/9405257

    work page arXiv 1994

  18. [18]

    G. P. Lepage, A New Algorithm for Adaptive Multidi- mensional Integration, J. Comput. Phys.27, 192 (1978)

    work page 1978

  19. [19]

    Ohl, Vegas revisited: Adaptive Monte Carlo integra- tion beyond factorization, Comput

    T. Ohl, Vegas revisited: Adaptive Monte Carlo integra- tion beyond factorization, Comput. Phys. Commun.120, 13 (1999), arXiv:hep-ph/9806432

    work page arXiv 1999

  20. [20]

    G. P. Lepage, Adaptive multidimensional integration: VEGAS enhanced, J. Comput. Phys.439, 110386 (2021), arXiv:2009.05112 [physics.comp-ph]

    work page arXiv 2021

  21. [22]

    Tabak and E

    E. Tabak and E. Vanden-Eijnden, Density estimation by dual ascent of the log-likelihood, Communications in Mathematical Sciences - COMMUN MATH SCI8 (2010)

    work page 2010

  22. [23]

    E. G. Tabak and C. V. Turner, A family of nonparametric density estimation algorithms, Communications on Pure and Applied Mathematics66, 145 (2013)

    work page 2013

  23. [24]

    L. Dinh, D. Krueger, and Y. Bengio, NICE: non-linear independent components estimation, in3rd Interna- tional Conference on Learning Representations, Work- shop Track Proceedings(2015) arXiv:1410.8516 [cs.LG]

    work page internal anchor Pith review arXiv 2015

  24. [25]

    M. D. Klimek and M. Perelstein, Neural Network-Based Approach to Phase Space Integration, SciPost Phys.9, 053 (2020), arXiv:1810.11509 [hep-ph]

    work page arXiv 2020

  25. [26]

    Bothmann, T

    E. Bothmann, T. Janßen, M. Knobbe, T. Schmale, and S. Schumann, Exploring phase space with Neu- ral Importance Sampling, SciPost Phys.8, 069 (2020), arXiv:2001.05478 [hep-ph]

    work page arXiv 2020

  26. [27]

    C. Gao, S. H¨ oche, J. Isaacson, C. Krause, and H. Schulz, Event Generation with Normalizing Flows, Phys. Rev. D 101, 076002 (2020), arXiv:2001.10028 [hep-ph]

    work page arXiv 2020

  27. [28]

    Heimel, R

    T. Heimel, R. Winterhalder, A. Butter, J. Isaacson, C. Krause, F. Maltoni, O. Mattelaer, and T. Plehn, Mad- NIS - Neural multi-channel importance sampling, SciPost Phys.15, 141 (2023), arXiv:2212.06172 [hep-ph]

    work page arXiv 2023

  28. [29]

    Verheyen, Event Generation and Density Estimation with Surjective Normalizing Flows, SciPost Phys.13, 047 (2022), arXiv:2205.01697 [hep-ph]

    R. Verheyen, Event Generation and Density Estimation with Surjective Normalizing Flows, SciPost Phys.13, 047 (2022), arXiv:2205.01697 [hep-ph]

    work page arXiv 2022

  29. [30]

    Heimel, N

    T. Heimel, N. Huetsch, F. Maltoni, O. Mattelaer, T. Plehn, and R. Winterhalder, The MadNIS reloaded, SciPost Phys.17, 023 (2024), arXiv:2311.01548 [hep-ph]

    work page arXiv 2024

  30. [31]

    Heimel, O

    T. Heimel, O. Mattelaer, T. Plehn, and R. Winter- halder, Differentiable MadNIS-Lite, SciPost Phys.18, 017 (2025), arXiv:2408.01486 [hep-ph]

    work page arXiv 2025

  31. [32]

    Badgeret al., Machine learning and LHC event gen- eration, SciPost Phys.14, 079 (2023), arXiv:2203.07460 [hep-ph]

    S. Badgeret al., Machine learning and LHC event gen- eration, SciPost Phys.14, 079 (2023), arXiv:2203.07460 [hep-ph]

    work page arXiv 2023

  32. [33]

    R. T. Q. Chen, Y. Rubanova, J. Bettencourt, and D. Duvenaud, Neural ordinary differential equations, inProceedings of the 32nd International Conference on Neural Information Processing Systems(Curran Asso- ciates Inc., Red Hook, NY, USA, 2018) p. 6572–6583, arXiv:1806.07366 [cs.LG]

    work page arXiv 2018

  33. [34]

    Flow Matching for Generative Modeling

    Y. Lipman, R. T. Q. Chen, H. Ben-Hamu, M. Nickel, and M. Le, Flow matching for generative modeling, in 7 The Eleventh International Conference on Learning Rep- resentations(2023) arXiv:2210.02747 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  34. [35]

    M. S. Albergo and E. Vanden-Eijnden, Building normal- izing flows with stochastic interpolants, inThe Eleventh International Conference on Learning Representations (2023) arXiv:2209.15571 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  35. [36]

    M. S. Albergo, N. M. Boffi, and E. Vanden-Eijnden, Stochastic interpolants: A unifying framework for flows and diffusions (2023), arXiv:2303.08797 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  36. [37]

    X. Liu, C. Gong, and Q. Liu, Flow straight and fast: Learning to generate and transfer data with rectified flow (2022), arXiv:2209.03003 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  37. [38]

    M¨ uller, B

    T. M¨ uller, B. Mcwilliams, F. Rousselle, M. Gross, and J. Nov´ ak, Neural importance sampling, ACM Trans. Graph.38(2019), arXiv:1808.03856 [cs.LG]

    work page arXiv 2019

  38. [39]

    Durkan, A

    C. Durkan, A. Bekasov, I. Murray, and G. Papamakar- ios, Neural spline flows, inAdvances in Neural Infor- mation Processing Systems, Vol. 32, edited by H. Wal- lach, H. Larochelle, A. Beygelzimer, F. d'Alch´ e-Buc, E. Fox, and R. Garnett (Curran Associates, Inc., 2019) arXiv:1906.04032 [stat.ML]

    work page arXiv 2019

  39. [40]

    LHAPDF6: parton density access in the LHC precision era

    A. Buckley, J. Ferrando, S. Lloyd, K. Nordstr¨ om, B. Page, M. R¨ ufenacht, M. Sch¨ onherr, and G. Watt, LHAPDF6: parton density access in the LHC precision era, Eur. Phys. J. C75, 132 (2015), arXiv:1412.7420 [hep- ph]

    work page Pith review arXiv 2015

  40. [41]

    Kullback and R

    S. Kullback and R. A. Leibler, On Information and Suf- ficiency, The Annals of Mathematical Statistics22, 79 (1951)

    work page 1951

  41. [42]

    A. Tong, K. FATRAS, N. Malkin, G. Huguet, Y. Zhang, J. Rector-Brooks, G. Wolf, and Y. Bengio, Improving and generalizing flow-based generative models with minibatch optimal transport, Trans. Mach. Learn. Res. (2024), arXiv:2302.00482 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  42. [43]

    Bothmann, T

    E. Bothmann, T. Childers, W. Giele, F. Herren, S. Hoeche, J. Isaacson, M. Knobbe, and R. Wang, Efficient phase-space generation for hadron collider event simulation, SciPost Phys.15, 169 (2023), arXiv:2302.10449 [hep-ph]

    work page arXiv 2023

  43. [44]

    Bothmann, T

    E. Bothmann, T. Childers, W. Giele, S. H¨ oche, J. Isaac- son, and M. Knobbe, A portable parton-level event gen- erator for the high-luminosity LHC, SciPost Phys.17, 081 (2024), arXiv:2311.06198 [hep-ph]

    work page arXiv 2024

  44. [45]

    Bothmann, W

    E. Bothmann, W. Giele, S. Hoeche, J. Isaacson, and M. Knobbe, Many-gluon tree amplitudes on modern GPUs: A case study for novel event generators, SciPost Phys. Codeb.2022, 3 (2022), arXiv:2106.06507 [hep-ph]

    work page arXiv 2022

  45. [46]

    Bothmann, J

    E. Bothmann, J. Isaacson, M. Knobbe, S. H¨ oche, and W. Giele, QCD tree amplitudes on modern GPUs: A case study for novel event generators, PoSICHEP2022, 222 (2022)

    work page 2022

  46. [47]

    Gleisberg, S

    T. Gleisberg, S. H¨ oche, F. Krauss, M. Sch¨ onherr, S. Schu- mann, F. Siegert, and J. Winter, Event generation with SHERPA 1.1, JHEP2009(02), 007, arXiv:0811.4622 [hep-ph]

    work page arXiv

  47. [48]

    Bothmann et al.,Event generation with Sherpa 2.2, SciPost Phys.7(2019) 034, arXiv:1905.09127 [hep-ph]

    E. Bothmannet al.(Sherpa), Event Generation with Sherpa 2.2, SciPost Phys.7, 034 (2019), arXiv:1905.09127 [hep-ph]

    work page arXiv 2019

  48. [49]

    Bothmannet al.(Sherpa), Event generation with Sherpa 3, JHEP2024(12), 156, arXiv:2410.22148 [hep- ph]

    E. Bothmannet al.(Sherpa), Event generation with Sherpa 3, JHEP2024(12), 156, arXiv:2410.22148 [hep- ph]

    work page arXiv

  49. [50]

    A comprehensive guide to the physics and usage of PYTHIA 8.3

    C. Bierlichet al., A comprehensive guide to the physics and usage of PYTHIA 8.3, SciPost Phys. Codeb.2022, 8 (2022), arXiv:2203.11601 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  50. [51]

    An Introduction to PYTHIA 8.2

    T. Sj¨ ostrand, S. Ask, J. R. Christiansen, R. Corke, N. De- sai, P. Ilten, S. Mrenna, S. Prestel, C. O. Rasmussen, and P. Z. Skands, An introduction to PYTHIA 8.2, Comput. Phys. Commun.191, 159 (2015), arXiv:1410.3012 [hep- ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  51. [52]

    Sjostrand, S

    T. Sjostrand, S. Mrenna, and P. Z. Skands, PYTHIA 6.4 Physics and Manual, JHEP2006(05), 026, arXiv:hep- ph/0603175

    work page arXiv

  52. [53]

    A Brief Introduction to PYTHIA 8.1

    T. Sjostrand, S. Mrenna, and P. Z. Skands, A Brief Intro- duction to PYTHIA 8.1, Comput. Phys. Commun.178, 852 (2008), arXiv:0710.3820 [hep-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2008

  53. [54]

    Melia, Dyck words and multiquark primitive ampli- tudes, Phys

    T. Melia, Dyck words and multiquark primitive ampli- tudes, Phys. Rev. D88, 014020 (2013), arXiv:1304.7809 [hep-ph]

    work page arXiv 2013

  54. [55]

    Melia, Dyck words and multi-quark amplitudes, PoS RADCOR2013, 031 (2013)

    T. Melia, Dyck words and multi-quark amplitudes, PoS RADCOR2013, 031 (2013)

    work page 2013

  55. [56]

    Johansson and A

    H. Johansson and A. Ochirov, Color-Kinematics Du- ality for QCD Amplitudes, JHEP2016(01), 170, arXiv:1507.00332 [hep-ph]

    work page arXiv

  56. [57]

    H¨ oche, S

    S. H¨ oche, S. Prestel, and H. Schulz, Simulation of Vector Boson Plus Many Jet Final States at the High Luminosity LHC, Phys. Rev.D100, 014024 (2019), arXiv:1905.05120 [hep-ph]

    work page arXiv 2019

  57. [58]

    Bothmann, T

    E. Bothmann, T. Childers, C. G¨ utschow, S. H¨ oche, P. Hovland, J. Isaacson, M. Knobbe, and R. Latham, Efficient precision simulation of processes with many-jet final states at the LHC, Phys. Rev. D109, 014013 (2024), arXiv:2309.13154 [hep-ph]

    work page arXiv 2024

  58. [59]

    R. D. Ballet al.(NNPDF), Parton distributions for the LHC Run II, JHEP2015(04), 040, arXiv:1410.8849 [hep-ph]

    work page Pith review arXiv

  59. [60]

    Z. Bern, L. J. Dixon, F. Febres Cordero, S. H¨ oche, H. Ita, D. A. Kosower, D. Maˆ ıtre, and K. J. Ozeren, Next-to- Leading OrderW+ 5-Jet Production at the LHC, Phys. Rev. D88, 014025 (2013), arXiv:1304.1253 [hep-ph]

    work page arXiv 2013

  60. [61]

    In: Advances in Neural Information Processing Systems, vol

    M. Tancik, P. P. Srinivasan, B. Mildenhall, S. Fridovich- Keil, N. Raghavan, U. Singhal, R. Ramamoorthi, J. T. Barron, and R. Ng, Fourier features let networks learn high frequency functions in low dimensional domains, inProceedings of the 34th International Conference on Neural Information Processing Systems(Curran Asso- ciates Inc., Red Hook, NY, USA, 2...

    work page arXiv 2020

  61. [62]

    Decoupled Weight Decay Regularization

    I. Loshchilov and F. Hutter, Decoupled weight decay reg- ularization, in7th International Conference on Learning Representations(2019) arXiv:1711.05101 [cs.LG]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  62. [63]

    Rehman, O

    D. Rehman, O. Davis, J. Lu, J. Tang, M. Bronstein, Y. Bengio, A. Tong, and A. J. Bose, Efficient regression- based training of normalizing flows for boltzmann gener- ators (2025), arXiv:2506.01158 [cs.LG]

    work page arXiv 2025