← back

CHESS: CHEbyshev pSeudo-Spectral transport for Feynman integral differential equations

📄 arXiv:2606.26691 · 📥 PDF · 2026-06-25 · hep-th

Authors: Yuanche Liu [arXiv · scholar] , Yang Zhang [arXiv · scholar]

🕰 Orloj analysis

8.1
Total score
8.5
Consistency
8.0
Quality
AD relevance

Tento článek představuje CHESS, balíček pro Wolfram Language určený k vysoce přesnému numerickému řešení diferenciálních rovnic pro Feynmanovy integrály pomocí Chebyshevovy pseudospektrální metody. Nástroj kombinuje spektrální kolokaci, řídkou maticovou montáž a regularizaci reziduí, přičemž benchmarky ukazují rychlou konvergenci a shodu s referenčními daty.

💡 Jde o cenný nový výpočetní nástroj, který nabízí praktická vylepšení v rychlosti a efektivitě pro složité výpočty Feynmanových integrálů v kvantové teorii pole.

Categories: PRT-3 MET-4 MET-1 QFD-6

✓ code_available, falsifiable, limit_reductions, modest_claims

⚠ Wolfram Language package may limit accessibility for some researchers, Wall time comparisons based on fixed local-series might not be exhaustive

📄 Abstract

We present CHESS (CHEbyshev pSeudo Spectrum), a Wolfram Language package for high-precision one-dimensional transport of ε-factorized differential equations for Feynman master integrals. The solver works with the matrix obtained by pulling a differential one-form to a chosen path. This matrix may be supplied directly, or assembled from constant matrices and precomputed scalar pullbacks of the one-forms. The program combines Chebyshev-Lobatto spectral collocation, sparse matrix assembly, sequential propagation in the ε-expansion, and residue-based regularization of spurious regular singular endpoints. Benchmarks for large multi-scale integral families show rapid node convergence and agreement with independent reference data where such data are available. In the fixed local-series comparison used here, the Chebyshev transports also give shorter wall times; the reported process-tree memory usage is comparable for the smaller parallel runs and lower for the largest benchmark system in that comparison.

📄 arXiv abstract page 📥 PDF