Machine Can Automatically Discover Parametric Functions to Model HEP Data

Abstract

In HEP data analyses, finding an adequate function to model binned data has largely relied on a manual process: guess a functional form by intuition, fit, examine, then repeat until successful. We show that this iterative process can be automated by a machine using symbolic regression, which performs a data-driven search over function space without requiring prior knowledge of what an adequate function should look like. We present the SymbolFit package, which pairs symbolic regression with uncertainty modeling to target HEP analysis use cases, and demonstrate it on the CMS and ATLAS Run 2 dijet spectra: 560 independent seeded runs across seven simple fit configurations generated over 1000 functions fitting the spectra with χ²/NDF≈ 1, and 111 of the runs rediscovered the very dijet and UA2 functions used in published dijet searches.

Ten-loop fixed-dimension renormalization-group functions of the three-dimensional O(N) model from distributed GPU-accelerated parametric integration
Paolo Ribeca
arXiv29 Oct 2026PreprintCC BY-NC-ND 4.0Full textComputational physics High-energy & particle physicsarXiv
Generation, Accumulation, and Utilization of Data in the Automatically Controlled End-Station of a Soft X-ray Beamline
Yasunobu Ando, Masashige Miyamamoto, Keita Hiromori +4
arXiv8 Oct 2026PreprintarXiv licenceFull textApplied physics & instrumentation Computational physicsarXiv
Compact and stable representation of real-frequency spectral functions for machine learning
Xinyang Dong, Ido Zemach, Lei Zhang +2
Physical Review Research3 Nov 2026Peer-reviewedCC BY 4.0Atomic, molecular & optical physicsCrossref
Physics-guided machine learning for predicting complete tool influence functions in flexible ball-end polishing
Shi Pan, Longxiang Li, Xingchang Li +6
Optics Express21 Oct 2026Peer-reviewedPublisher termsOptics & photonicsCrossref
How AI Agents Discover Scientific Equations: From Hydrotope Rediscovery to New Water-Wave Amplitudes
Zihan Zhou, Digvijay Wadekar, Matias Zaldarriaga
arXiv1 Oct 2026PreprintCC BY 4.0Full textHigh-energy & particle physicsarXiv

Cite

BibTeX RIS

BibTeX

@misc{tsoi2026machine,
  title = {{Machine Can Automatically Discover Parametric Functions to Model HEP Data}},
  author = {Ho Fung Tsoi and Dylan Rankin and Cecile Caillol and Miles Cranmer and Sridhara Dasu and Javier Duarte and Philip Harris and Elliot Lipeles},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2607.19750},
  eprint = {2607.19750},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2607.19750},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Machine Can Automatically Discover Parametric Functions to Model HEP Data
AU  - Ho Fung Tsoi
AU  - Dylan Rankin
AU  - Cecile Caillol
AU  - Miles Cranmer
AU  - Sridhara Dasu
AU  - Javier Duarte
AU  - Philip Harris
AU  - Elliot Lipeles
PY  - 2026
DA  - 2026/10/19
PB  - arXiv
DO  - 10.48550/arxiv.2607.19750
UR  - https://arxiv.org/abs/2607.19750
AB  - In HEP data analyses, finding an adequate function to model binned data has largely relied on a manual process: guess a functional form by intuition, fit, examine, then repeat until successful. We show that this iterative process can be automated by a machine using symbolic regression, which performs a data-driven search over function space without requiring prior knowledge of what an adequate function should look like. We present the SymbolFit package, which pairs symbolic regression with uncertainty modeling to target HEP analysis use cases, and demonstrate it on the CMS and ATLAS Run 2 dijet spectra: 560 independent seeded runs across seven simple fit configurations generated over 1000 functions fitting the spectra with χ^2/NDF≈ 1, and 111 of the runs rediscovered the very dijet and UA2 functions used in published dijet searches.
ER  -

Abstract

Related

Cite