pith. sign in

arxiv: 2606.10698 · v1 · pith:X42VR4T5new · submitted 2026-06-09 · ✦ hep-ph · cs.LG· hep-th

Efficient AI-Inspired Reduction of Feynman Integrals via Tube Seeding

Justin Berman , Francois Charton , Andres Luna , Matthias Wilhelm , Mao Zeng This is my paper

Pith reviewed 2026-06-27 12:27 UTC · model grok-4.3

classification ✦ hep-ph cs.LGhep-th
keywords Feynman integralsIntegration-by-parts reductionLaporta algorithmMachine learningSeeding strategyMulti-loop calculationsTube seeding
0
0 comments X

The pith

A tube-shaped selection of seed integrals lets the standard Laporta algorithm reduce high-numerator-power Feynman integrals with only linear growth in the number of seeds.

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

The paper shows that machine learning can identify a sparse seeding strategy for integration-by-parts reduction in which the required seed integrals lie inside a thin tube-like region that follows a zigzag path from the target integral to the master integrals. This replaces the usual polynomial growth in seed count with linear growth in the numerator power, while still using the unmodified Laporta algorithm. The approach is demonstrated by successfully reducing non-planar two-loop five-point integrals of rank 20 and complete sets of rank-10 integrals, tasks that exceed the reach of conventional seeding on the same hardware. The method therefore removes a practical bottleneck that has limited the complexity of multi-loop calculations in particle physics.

Core claim

Restricting the seed integrals for Laporta reduction to a thin tube-like region around a zigzag path from the target integral to the master integrals yields a complete reduction whose seed count scales linearly with numerator power rather than polynomially.

What carries the argument

The thin tube-like region of seed integrals around a zigzag path that connects the target integral to the master integrals.

If this is right

  • Non-planar two-loop five-point integrals of rank 20 become reducible over finite fields with modest resources.
  • Complete families of top-level rank-10 integrals can be reduced in separate chunks with substantially lower time and memory than existing methods.
  • The same linear scaling applies to integrals with large numerator powers that appear in phenomenological calculations.
  • The strategy remains compatible with the unmodified Laporta algorithm and existing reduction codes.

Where Pith is reading between the lines

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

  • If a systematic way to construct the zigzag path is found, the method could extend to three-loop or higher integrals without further machine-learning searches.
  • Dividing large families into independent chunks, as shown for rank-10 sets, may become a standard preprocessing step for amplitude-level reductions.
  • The linear seed growth could make automated reduction of entire two-loop amplitudes routine on desktop hardware.

Load-bearing premise

The thin tube around the zigzag path always contains every seed integral needed for a complete reduction without missing relations.

What would settle it

A concrete high-rank integral for which the reduction performed with only the tube seeds produces incorrect coefficients for the master integrals or leaves unresolved relations.

read the original abstract

In this paper, we use machine learning to discover a new seeding strategy for integration-by-parts reduction of Feynman integrals, which is a frequent bottleneck in state-of-the-art calculations in theoretical particle and gravitational-wave physics. Our strategy allows us to reduce multi-loop integrals with large numerator powers via essentially the standard Laporta algorithm but with a sparse selection of seed integrals that grows only linearly with the numerator power, whereas existing strategies lead to growth with a polynomial power that increases with the complexity of the integral being reduced. The seeds are restricted to a thin tube-like region that connects the target integral to the master integrals along a zigzag path. We demonstrate the power of our approach by reducing non-planar 2-loop 5-point integrals of rank 20 with numerical kinematics over a finite field, which is prohibitively difficult for the Laporta algorithm with conventional seeding. Going beyond individual integrals, we further demonstrate the reduction of a complete set of top-level rank-10 integrals by dividing the target integrals into several chunks, each of which can be solved by our sparse seeding strategy with considerably less time and a significantly lower memory footprint than other state-of-the-art strategies, making the approach well-suited for phenomenological applications. We provide a proof-of-principle implementation on GitHub at https://github.com/andreslunagodoy/tube_seeding.

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 paper claims that machine learning can discover a sparse 'tube seeding' strategy for Laporta-style IBP reduction of Feynman integrals. By restricting seed integrals to a thin tube-like region around an ML-discovered zigzag path from target to masters, the number of required seeds grows only linearly with numerator rank, in contrast to the polynomial growth of conventional seeding. The approach is demonstrated by reducing non-planar two-loop five-point integrals of rank 20 and complete rank-10 families over finite fields, with a proof-of-principle GitHub implementation provided.

Significance. If the tube method proves reliable beyond the tested cases, it would materially lower the computational barrier for high-rank multi-loop reductions that currently limit phenomenological calculations in hep-ph and gravitational-wave physics. The open-source code is a concrete strength that enables direct verification and extension.

major comments (2)
  1. [Abstract and method description] The central claim (abstract and § on strategy) that restricting seeds to the thin tube yields a complete reduction rests on the unproven assumption that the zigzag path captures all necessary IBP relations. The manuscript supplies only empirical success on two specific topologies (non-planar 2-loop 5-point rank 20 and rank-10 families); no theorem, syzygy argument, or exhaustive check establishes that the tube is closed under the IBP ideal for arbitrary graphs or higher ranks.
  2. [Results on rank-20 integrals] No quantitative error analysis, completeness metric, or failure-mode study is reported for the finite-field reductions. The abstract states only qualitative feasibility; without reported counts of missed relations, residual rank, or comparison of generated basis size against a known complete reduction, it is impossible to assess whether the obtained reductions are exact.
minor comments (2)
  1. [Abstract] The abstract and introduction would benefit from explicit statements of the tested topologies, the precise definition of 'tube width', and the finite-field characteristic used, to allow immediate reproducibility from the GitHub repository.
  2. [Method and figures] Figure captions and the description of the ML discovery procedure should clarify whether the zigzag path is found once per topology or per integral family, and how the tube radius is chosen.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive report and for recognizing the potential significance of the tube-seeding approach. We address the two major comments point by point below, acknowledging the empirical character of the method while outlining specific revisions to improve clarity and documentation.

read point-by-point responses

  1. Referee: [Abstract and method description] The central claim (abstract and § on strategy) that restricting seeds to the thin tube yields a complete reduction rests on the unproven assumption that the zigzag path captures all necessary IBP relations. The manuscript supplies only empirical success on two specific topologies (non-planar 2-loop 5-point rank 20 and rank-10 families); no theorem, syzygy argument, or exhaustive check establishes that the tube is closed under the IBP ideal for arbitrary graphs or higher ranks.

    Authors: We agree that the completeness of the reduction is established empirically rather than by a general theorem or syzygy argument. The zigzag path and associated tube were obtained via machine-learning search and then validated by explicit reduction of the two cited topologies, including cases where standard Laporta seeding becomes intractable. The manuscript presents the method as a practical, ML-discovered heuristic rather than a universally proven algorithm. In the revised version we will add an explicit statement in the abstract and in the strategy section clarifying the empirical basis and the current scope of validation. revision: partial

  2. Referee: [Results on rank-20 integrals] No quantitative error analysis, completeness metric, or failure-mode study is reported for the finite-field reductions. The abstract states only qualitative feasibility; without reported counts of missed relations, residual rank, or comparison of generated basis size against a known complete reduction, it is impossible to assess whether the obtained reductions are exact.

    Authors: The referee correctly notes that the original manuscript reports only qualitative success. In the revision we will augment the results section with quantitative metrics for the rank-20 example: the number of seed integrals, the size of the linear system, the final master-integral basis dimension, and explicit verification that every target integral reduces to zero residual. We will also include a short comparison against lower-rank reductions (where independent checks exist) and a brief discussion of observed failure modes. Because all arithmetic is performed over finite fields, exactness is guaranteed once the linear system is solved; the added numbers will make this verification transparent. revision: yes

Circularity Check

0 steps flagged

No significant circularity; algorithmic claim rests on external benchmarks

full rationale

The paper's central result is an empirically validated seeding heuristic discovered via ML and then applied within the standard Laporta algorithm. Reductions of rank-20 non-planar two-loop five-point integrals and chunked rank-10 sets are shown to succeed where conventional polynomial-growth seeding fails, with explicit finite-field timings and a public GitHub implementation. No equation equates a derived quantity to a fitted parameter by construction, no self-citation supplies a load-bearing uniqueness theorem, and the tube-completeness claim is presented as an observed property of the tested topologies rather than a self-referential definition. The derivation chain therefore remains self-contained against the supplied numerical evidence.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work introduces an algorithmic strategy without new physical axioms or fitted parameters; it relies on standard properties of integration-by-parts identities and finite-field arithmetic.

axioms (1)
  • standard math Integration-by-parts identities generate a linear system whose solution space is spanned by a finite set of master integrals.
    Invoked implicitly as the foundation of the Laporta algorithm throughout the abstract.

pith-pipeline@v0.9.1-grok · 5775 in / 1321 out tokens · 19047 ms · 2026-06-27T12:27:05.869790+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

90 extracted references · 3 canonical work pages

  1. [1]

    Tkachov,A theorem on analytical calculability of 4-loop renormalization group functions,Phys

    F.V. Tkachov,A theorem on analytical calculability of 4-loop renormalization group functions,Phys. Lett. B100(1981) 65

    1981

  2. [2]

    Chetyrkin and F.V

    K.G. Chetyrkin and F.V. Tkachov,Integration by Parts: The Algorithm to Calculate beta Functions in 4 Loops,Nucl. Phys. B192(1981) 159

    1981

  3. [3]

    Laporta,High precision calculation of multiloop Feynman integrals by difference equations,Int

    S. Laporta,High precision calculation of multiloop Feynman integrals by difference equations,Int. J. Mod. Phys. A15(2000) 5087 [hep-ph/0102033]

    Pith/arXiv arXiv 2000

  4. [4]

    Grozin,Integration by parts: An Introduction,Int

    A.G. Grozin,Integration by parts: An Introduction,Int. J. Mod. Phys. A26(2011) 2807 [1104.3993]

    Pith/arXiv arXiv 2011

  5. [5]

    Smirnov,Analytic tools for Feynman integrals, Springer, Berlin, Heidelberg (2012), 10.1007/978-3-642-34886-0

    V.A. Smirnov,Analytic tools for Feynman integrals, vol. 250, Springer Berlin, Heidelberg (2012), 10.1007/978-3-642-34886-0

    work page doi:10.1007/978-3-642-34886-0 2012

  6. [6]

    Weinzierl,Feynman Integrals

    S. Weinzierl,Feynman Integrals. A Comprehensive Treatment for Students and Researchers, UNITEXT for Physics, Springer (2022), 10.1007/978-3-030-99558-4, [2201.03593]

    work page doi:10.1007/978-3-030-99558-4 2022

  7. [7]

    Kotikov,Differential equations method: New technique for massive Feynman diagrams calculation,Phys

    A.V. Kotikov,Differential equations method: New technique for massive Feynman diagrams calculation,Phys. Lett. B254(1991) 158

    1991

  8. [8]

    Gehrmann and E

    T. Gehrmann and E. Remiddi,Differential equations for two loop four point functions,Nucl. Phys. B580(2000) 485 [hep-ph/9912329]

    Pith/arXiv arXiv 2000

  9. [9]

    Henn,Multiloop Integrals in Dimensional Regularization Made Simple,Phys

    J.M. Henn,Multiloop Integrals in Dimensional Regularization Made Simple,Phys. Rev. Lett.110(2013) 251601 [1304.1806]

    Pith/arXiv arXiv 2013

  10. [10]

    Anastasiou and A

    C. Anastasiou and A. Lazopoulos,Automatic integral reduction for higher order perturbative calculations,JHEP07(2004) 046 [hep-ph/0404258]

    Pith/arXiv arXiv 2004

  11. [11]

    Smirnov,Algorithm FIRE – Feynman Integral REduction,JHEP10(2008) 107 [0807.3243]

    A.V. Smirnov,Algorithm FIRE – Feynman Integral REduction,JHEP10(2008) 107 [0807.3243]

    Pith/arXiv arXiv 2008

  12. [12]

    Smirnov and M

    A.V. Smirnov and M. Zeng,FIRE 6.5: Feynman integral reduction with new simplification library,Comput. Phys. Commun.302(2024) 109261 [2311.02370]

    arXiv 2024

  13. [13]

    Smirnov and M

    A.V. Smirnov and M. Zeng,FIRE 7: Automatic Reduction with Modular Approach, 2510.07150

    arXiv

  14. [14]

    von Manteuffel and C

    A. von Manteuffel and C. Studerus,Reduze 2 - Distributed Feynman Integral Reduction,1201.4330

    Pith/arXiv arXiv

  15. [15]

    Lee,Presenting LiteRed: a tool for the Loop InTEgrals REDuction,1212.2685

    R.N. Lee,Presenting LiteRed: a tool for the Loop InTEgrals REDuction,1212.2685

    Pith/arXiv arXiv

  16. [16]

    Maierh¨ ofer, J

    P. Maierh¨ ofer, J. Usovitsch and P. Uwer,Kira—A Feynman integral reduction program,Comput. Phys. Commun.230(2018) 99 [1705.05610]

    Pith/arXiv arXiv 2018

  17. [17]

    Klappert, F

    J. Klappert, F. Lange, P. Maierh¨ ofer and J. Usovitsch,Integral reduction with Kira 2.0 and finite field methods,Comput. Phys. Commun.266(2021) 108024 [2008.06494]. – 55 –

    Pith/arXiv arXiv 2021

  18. [18]

    Lange, J

    F. Lange, J. Usovitsch and Z. Wu,Kira 3: integral reduction with efficient seeding and optimized equation selection,Comput. Phys. Commun.322(2026) 109999 [2505.20197]

    arXiv 2026

  19. [19]

    Peraro,FiniteFlow: multivariate functional reconstruction using finite fields and dataflow graphs,JHEP07(2019) 031 [1905.08019]

    T. Peraro,FiniteFlow: multivariate functional reconstruction using finite fields and dataflow graphs,JHEP07(2019) 031 [1905.08019]

    Pith/arXiv arXiv 2019

  20. [20]

    X. Guan, X. Liu, Y.-Q. Ma and W.-H. Wu,Blade: A package for block-triangular form improved Feynman integrals decomposition,Comput. Phys. Commun.310 (2025) 109538 [2405.14621]

    arXiv 2025

  21. [21]

    Improved integral reduction with kira

    J. Usovitsch, “Improved integral reduction with kira.” Talk at QCD meets Gravity at CERN, December 13th 2023. Slides available at https://indico.cern.ch/event/1317494/contributions/5697745/ attachments/2770593/4827307/Kira_QCD_meets_Gravity.pdf

    arXiv 2023

  22. [22]

    Driesse, G.U

    M. Driesse, G.U. Jakobsen, G. Mogull, J. Plefka, B. Sauer and J. Usovitsch, Conservative Black Hole Scattering at Fifth Post-Minkowskian and First Self-Force Order,Phys. Rev. Lett.132(2024) 241402 [2403.07781]

    arXiv 2024

  23. [23]

    Z. Bern, E. Herrmann, R. Roiban, M.S. Ruf, A.V. Smirnov, V.A. Smirnov et al., Amplitudes, supersymmetric black hole scattering atO G5 , and loop integration, JHEP10(2024) 023 [2406.01554]

    arXiv 2024

  24. [24]

    Gluza, K

    J. Gluza, K. Kajda and D.A. Kosower,Towards a Basis for Planar Two-Loop Integrals,Phys. Rev. D83(2011) 045012 [1009.0472]

    Pith/arXiv arXiv 2011

  25. [25]

    Larsen and Y

    K.J. Larsen and Y. Zhang,Integration-by-parts reductions from unitarity cuts and algebraic geometry,Phys. Rev. D93(2016) 041701 [1511.01071]

    Pith/arXiv arXiv 2016

  26. [26]

    Boehm, D

    J. Boehm, D. Bendle, W. Decker, A. Georgoudis, F.-J. Pfreundt, M. Rahn et al., Module intersection for the integration-by-parts reduction of multi-loop feynman integrals,PoSMA2019(2022) 004 [2010.06895]

    arXiv 2022

  27. [27]

    Z. Wu, J. Boehm, R. Ma, H. Xu and Y. Zhang,NeatIBP 1.0, a package generating small-size integration-by-parts relations for Feynman integrals,Comput. Phys. Commun.295(2024) 108999 [2305.08783]

    arXiv 2024

  28. [28]

    Z. Wu, J. B¨ ohm, R. Ma, J. Usovitsch, Y. Xu and Y. Zhang,Performing integration-by-parts reductions using NeatIBP 1.1 + Kira,Comput. Phys. Commun. 316(2025) 109798 [2502.20778]

    arXiv 2025

  29. [29]

    Lee,Group structure of the integration-by-part identities and its application to the reduction of multiloop integrals,JHEP07(2008) 031 [0804.3008]

    R.N. Lee,Group structure of the integration-by-part identities and its application to the reduction of multiloop integrals,JHEP07(2008) 031 [0804.3008]

    Pith/arXiv arXiv 2008

  30. [30]

    Mastrolia and S

    P. Mastrolia and S. Mizera,Feynman Integrals and Intersection Theory,JHEP02 (2019) 139 [1810.03818]

    Pith/arXiv arXiv 2019

  31. [31]

    Frellesvig, F

    H. Frellesvig, F. Gasparotto, M.K. Mandal, P. Mastrolia, L. Mattiazzi and S. Mizera,Vector Space of Feynman Integrals and Multivariate Intersection Numbers,Phys. Rev. Lett.123(2019) 201602 [1907.02000]. – 56 –

    arXiv 2019

  32. [32]

    von Hippel and M

    M. von Hippel and M. Wilhelm,Refining Integration-by-Parts Reduction of Feynman Integrals with Machine Learning,JHEP05(2025) 185 [2502.05121]

    arXiv 2025

  33. [33]

    Song, T.-Z

    Z.-Y. Song, T.-Z. Yang, Q.-H. Cao, M.-x. Luo and H.X. Zhu,Explainable AI-assisted Optimization for Feynman Integral Reduction,2502.09544

    arXiv

  34. [34]

    Zeng,Reinforcement learning and metaheuristics for Feynman-integral reduction, Phys

    M. Zeng,Reinforcement learning and metaheuristics for Feynman-integral reduction, Phys. Rev. D112(2025) 114051 [2504.16045]

    arXiv 2025

  35. [35]

    Shih,Learning to Unscramble Feynman Loop Integrals with SAILIR,2604.05034

    D. Shih,Learning to Unscramble Feynman Loop Integrals with SAILIR,2604.05034

    Pith/arXiv arXiv

  36. [36]

    Smith and M

    S. Smith and M. Zeng,Feynman integral reduction using syzygy-constrained symbolic reduction rules,JHEP01(2026) 102 [2507.11140]

    arXiv 2026

  37. [37]

    Liu and A

    J.W. Liu and A. Mitov,Untangling the IBP Equations,2512.05923

    arXiv

  38. [38]

    de la Cruz and D.A

    L. de la Cruz and D.A. Kosower,Seedless Reduction of Feynman Integrals, 2602.22111

    arXiv

  39. [39]

    von Gersdorff and V

    G. von Gersdorff and V. Lessa,Feynman integral reduction by covariant differentiation,2604.09810

    Pith/arXiv arXiv

  40. [40]

    Feng,Generation function for one-loop tensor reduction,Commun

    B. Feng,Generation function for one-loop tensor reduction,Commun. Theor. Phys. 75(2023) 025203 [2209.09517]

    arXiv 2023

  41. [41]

    C. Hu, T. Li, J. Shen and Y. Xu,An explicit expression of generating function for one-loop tensor reduction,JHEP02(2024) 158 [2308.13336]

    arXiv 2024

  42. [42]

    X. Guan, X. Li and Y.-Q. Ma,Exploring the linear space of feynman integrals via generating functions,Phys. Rev. D108(2023) 034027 [2306.02927]

    arXiv 2023

  43. [43]

    T. Li, Y. Song and L. Zhang,Solving arbitrary one-loop reduction via generating function,Eur. Phys. J. C85(2025) 142 [2404.04644]

    arXiv 2025

  44. [44]

    C. Hu, B. Feng, J. Shen and Y. Zhang,General one-loop generating function by integration-by-part relations,Phys. Rev. D111(2025) 076026 [2403.16040]

    arXiv 2025

  45. [45]

    X. Chen, B. Feng and L. Zhang,Tensor Reduction of Sunset by Generating Function,2509.18730

    arXiv

  46. [46]

    Hu, W.-D

    C. Hu, W.-D. Li and X. Li,Generating function of loop reduction by baikov representation,2504.02573

    arXiv

  47. [47]

    B. Feng, X. Li, Y. Liu, Y.-Q. Ma and Y. Zhang,Symbolic Reduction of Multi-loop Feynman Integrals via Generating Functions,2509.21769

    arXiv

  48. [48]

    Brunello, V

    G. Brunello, V. Chestnov and P. Mastrolia,Intersection numbers from companion tensor algebra,JHEP07(2025) 045 [2408.16668]

    arXiv 2025

  49. [49]

    Kant,Finding linear dependencies in integration-by-parts equations: A Monte Carlo approach,Comput

    P. Kant,Finding linear dependencies in integration-by-parts equations: A Monte Carlo approach,Comput. Phys. Commun.185(2014) 1473 [1309.7287]

    Pith/arXiv arXiv 2014

  50. [50]

    von Manteuffel and R.M

    A. von Manteuffel and R.M. Schabinger,A novel approach to integration by parts reduction,Phys. Lett. B744(2015) 101 [1406.4513]. – 57 –

    Pith/arXiv arXiv 2015

  51. [51]

    Peraro,Scattering amplitudes over finite fields and multivariate functional reconstruction,JHEP12(2016) 030 [1608.01902]

    T. Peraro,Scattering amplitudes over finite fields and multivariate functional reconstruction,JHEP12(2016) 030 [1608.01902]

    Pith/arXiv arXiv 2016

  52. [52]

    Abreu, J

    S. Abreu, J. Dormans, F. Febres Cordero, H. Ita and B. Page,Analytic Form of Planar Two-Loop Five-Gluon Scattering Amplitudes in QCD,Phys. Rev. Lett.122 (2019) 082002 [1812.04586]

    Pith/arXiv arXiv 2019

  53. [53]

    Klappert and F

    J. Klappert and F. Lange,Reconstructing rational functions with FireFly,Comput. Phys. Commun.247(2020) 106951 [1904.00009]

    arXiv 2020

  54. [54]

    Laurentis and D

    G. Laurentis and D. Maˆ ıtre,Extracting analytical one-loop amplitudes from numerical evaluations,JHEP07(2019) 123 [1904.04067]

    arXiv 2019

  55. [55]

    De Laurentis and B

    G. De Laurentis and B. Page,Ans¨ atze for scattering amplitudes from p-adic numbers and algebraic geometry,JHEP12(2022) 140 [2203.04269]

    arXiv 2022

  56. [56]

    Magerya,Rational Tracer: a Tool for Faster Rational Function Reconstruction, 2211.03572

    V. Magerya,Rational Tracer: a Tool for Faster Rational Function Reconstruction, 2211.03572

    arXiv

  57. [57]

    Belitsky, A.V

    A.V. Belitsky, A.V. Smirnov and R.V. Yakovlev,Balancing act: Multivariate rational reconstruction for IBP,Nucl. Phys. B993(2023) 116253 [2303.02511]

    arXiv 2023

  58. [58]

    Chawdhry,p-adic reconstruction of rational functions in multiloop amplitudes, Phys

    H.A. Chawdhry,p-adic reconstruction of rational functions in multiloop amplitudes, Phys. Rev. D110(2024) 056028 [2312.03672]

    arXiv 2024

  59. [59]

    Liu,Reconstruction of rational functions made simple,Phys

    X. Liu,Reconstruction of rational functions made simple,Phys. Lett. B850(2024) 138491 [2306.12262]

    arXiv 2024

  60. [60]

    Maier,Scaling up to Multivariate Rational Function Reconstruction,Comput

    A. Maier,Scaling up to Multivariate Rational Function Reconstruction,Comput. Phys. Commun.317(2025) 109827 [2409.08757]

    arXiv 2025

  61. [61]

    Smirnov and A.V

    A.V. Smirnov and A.V. Petukhov,The Number of Master Integrals is Finite,Lett. Math. Phys.97(2011) 37 [1004.4199]

    Pith/arXiv arXiv 2011

  62. [62]

    Machine learning for feynman integral reduction

    M. Zeng, “Machine learning for feynman integral reduction.” Talk at LoopFest XXIII Conference at the University of Alberta, May 23th 2025. Slides available at https://indico.cern.ch/event/1446102/contributions/6462127/ attachments/3073832/5438788/talkLoopFest.pdf

    arXiv 2025

  63. [63]

    Schulman, F

    J. Schulman, F. Wolski, P. Dhariwal, A. Radford and O. Klimov,Proximal Policy Optimization Algorithms,1707.06347

    Pith/arXiv arXiv

  64. [64]

    Towers, A

    M. Towers, A. Kwiatkowski, J. Terry, J.U. Balis, G.D. Cola, T. Deleu et al., Gymnasium: A standard interface for reinforcement learning environments,arXiv preprint arXiv:2407.17032(2024) [2407.17032]

    Pith/arXiv arXiv 2024

  65. [65]

    Huang, R.F.J

    S. Huang, R.F.J. Dossa, C. Ye, J. Braga, D. Chakraborty, K. Mehta et al.,Cleanrl: High-quality single-file implementations of deep reinforcement learning algorithms, Journal of Machine Learning Research23(2022) 1

    2022

  66. [66]

    Hansen and A

    N. Hansen and A. Ostermeier,Completely derandomized self-adaptation in evolution strategies,Evolutionary Computation9(2001) 159. – 58 –

    2001

  67. [67]

    CMA-ES/pycma on Github

    N. Hansen, Y. Akimoto and P. Baudis, “CMA-ES/pycma on Github.” Zenodo, DOI:10.5281/zenodo.2559634, Feb., 2019. 10.5281/zenodo.2559634

    work page doi:10.5281/zenodo.2559634 2019

  68. [68]

    Hendrycks and K

    D. Hendrycks and K. Gimpel,Gaussian Error Linear Units (GELUs),1606.08415

    Pith/arXiv arXiv

  69. [69]

    Kingma and J

    D.P. Kingma and J. Ba,Adam: A Method for Stochastic Optimization, in International Conference on Learning Representations, 12, 2014 [1412.6980]

    Pith/arXiv arXiv 2014

  70. [70]

    Abreu, L.J

    S. Abreu, L.J. Dixon, E. Herrmann, B. Page and M. Zeng,The two-loop five-point amplitude inN= 4super-Yang-Mills theory,Phys. Rev. Lett.122(2019) 121603 [1812.08941]

    Pith/arXiv arXiv 2019

  71. [71]

    Chicherin, T

    D. Chicherin, T. Gehrmann, J.M. Henn, P. Wasser, Y. Zhang and S. Zoia,Analytic result for a two-loop five-particle amplitude,Phys. Rev. Lett.122(2019) 121602 [1812.11057]

    Pith/arXiv arXiv 2019

  72. [72]

    Abreu, L.J

    S. Abreu, L.J. Dixon, E. Herrmann, B. Page and M. Zeng,The two-loop five-point amplitude inN= 8supergravity,JHEP03(2019) 123 [1901.08563]

    Pith/arXiv arXiv 2019

  73. [73]

    Chicherin, T

    D. Chicherin, T. Gehrmann, J.M. Henn, P. Wasser, Y. Zhang and S. Zoia,The two-loop five-particle amplitude inN= 8supergravity,JHEP03(2019) 115 [1901.05932]

    Pith/arXiv arXiv 2019

  74. [74]

    Badger, C

    S. Badger, C. Brønnum-Hansen, H.B. Hartanto and T. Peraro,Analytic helicity amplitudes for two-loop five-gluon scattering: the single-minus case,JHEP01(2019) 186 [1811.11699]

    Pith/arXiv arXiv 2019

  75. [75]

    Abreu, J

    S. Abreu, J. Dormans, F. Febres Cordero, H. Ita, B. Page and V. Sotnikov,Analytic Form of the Planar Two-Loop Five-Parton Scattering Amplitudes in QCD,JHEP 05(2019) 084 [1904.00945]

    Pith/arXiv arXiv 2019

  76. [76]

    Badger, D

    S. Badger, D. Chicherin, T. Gehrmann, G. Heinrich, J.M. Henn, T. Peraro et al., Analytic form of the full two-loop five-gluon all-plus helicity amplitude,Phys. Rev. Lett.123(2019) 071601 [1905.03733]

    arXiv 2019

  77. [77]

    Agarwal, F

    B. Agarwal, F. Buccioni, F. Devoto, G. Gambuti, A. von Manteuffel and L. Tancredi,Five-parton scattering in QCD at two loops,Phys. Rev. D109(2024) 094025 [2311.09870]

    arXiv 2024

  78. [78]

    Bendle, J

    D. Bendle, J. B¨ ohm, W. Decker, A. Georgoudis, F.-J. Pfreundt, M. Rahn et al., Integration-by-parts reductions of Feynman integrals using Singular and GPI-Space, JHEP02(2020) 079 [1908.04301]

    arXiv 2020

  79. [79]

    Davis,Direct Methods for Sparse Linear Systems, SIAM, Philadelphia, PA (2006)

    T.A. Davis,Direct Methods for Sparse Linear Systems, SIAM, Philadelphia, PA (2006)

    2006

  80. [80]

    Duff, A.M

    I.S. Duff, A.M. Erisman and J.K. Reid,Direct Methods for Sparse Matrices, Oxford University Press, Oxford (1986)

    1986

Showing first 80 references.