Explainable AI-assisted optimization for Feynman integral reduction
4 affiliations
- 1Peking University
- 2South China Normal University
- 3Universität Zürich
- 4Beijing Computational Science Research Center
Abstract
A bstract We present a novel approach to optimizing the reduction of Feynman integrals using integration-by-parts identities. By developing a priority function through the FunSearch algorithm, which combines large language models and genetic algorithms, we achieve significant improvements in memory usage and computational efficiency compared to traditional methods. Our approach demonstrates substantial reductions in the required seeding integrals, making previously intractable integrals more manageable. Tested on a variety of Feynman integrals, including one-loop and multi-loop cases with planar and non-planar configurations, our method demonstrates remarkable scalability and adaptability. For reductions of certain Feynman integrals with many dots and numerators, we observed an improvement of several orders of magnitude over traditional methods. This work provides a powerful and interpretable framework for optimizing IBP reductions, paving the way for more efficient and practical calculations in high-energy physics.
Related
- Machine learning-assisted design and explainable optimization of CdSnP2-based integrated solar-photodetector devices
- Two Feynman integral families for vector boson fusion at NNLO QCD
- Reduction technique for expanding the Feynman diagrams in AdS2
- SecRed: A program for sector-wise IBP reduction of Feynman integrals
- Simultaneous Reduction of Observables and Measured Qudits in Entanglement-Assisted Quantum Local Recovery
BibTeX
@article{song2026explainable,
title = {{Explainable AI-assisted optimization for Feynman integral reduction}},
author = {Zhe Song and Tong-Zhi Yang and Qing-Hong Cao and Ming-xing Luo and Hua Xing Zhu},
journal = {Journal of High Energy Physics},
year = {2026},
month = jun,
doi = {10.1007/jhep06(2026)225},
url = {https://doi.org/10.1007/jhep06(2026)225},
note = {CC BY},
}
RIS
TY - JOUR
TI - Explainable AI-assisted optimization for Feynman integral reduction
AU - Zhe Song
AU - Tong-Zhi Yang
AU - Qing-Hong Cao
AU - Ming-xing Luo
AU - Hua Xing Zhu
PY - 2026
DA - 2026/06/22
JO - Journal of High Energy Physics
DO - 10.1007/jhep06(2026)225
UR - https://doi.org/10.1007/jhep06(2026)225
AB - A bstract We present a novel approach to optimizing the reduction of Feynman integrals using integration-by-parts identities. By developing a priority function through the FunSearch algorithm, which combines large language models and genetic algorithms, we achieve significant improvements in memory usage and computational efficiency compared to traditional methods. Our approach demonstrates substantial reductions in the required seeding integrals, making previously intractable integrals more manageable. Tested on a variety of Feynman integrals, including one-loop and multi-loop cases with planar and non-planar configurations, our method demonstrates remarkable scalability and adaptability. For reductions of certain Feynman integrals with many dots and numerators, we observed an improvement of several orders of magnitude over traditional methods. This work provides a powerful and interpretable framework for optimizing IBP reductions, paving the way for more efficient and practical calculations in high-energy physics.
SN - 1029-8479
SN - 1126-6708
SN - 1127-2236
ER -