Brilliance: toward autonomous computational research in particle accelerator physics
Abstract
Computational research in physics requires continuous interplay between simulation and scientific interpretation, demanding substantial specialist effort. Here we introduce Brilliance, a research agent that combines large language models, reusable physics knowledge and numerical simulation to conduct adaptive investigations in accelerator physics. Given a single user prompt, Brilliance designs a fourth-generation synchrotron light source that meets coupled emittance and nonlinear-acceptance requirements. Further exploration yields a design combining sub-10-pm emittance with a large dynamic aperture under realistic engineering constraints, suggesting a route towards hard-X-ray diffraction-limited operation, although the momentum-acceptance target remains unmet. Brilliance also calibrates a model of an operating light source against measured responses, accurately inferring its lattice and calibration errors. Across these studies, Brilliance autonomously revises its computational strategies in response to numerical evidence. These results demonstrate a route towards delegating sustained scientific investigations to agents equipped with domain expertise and numerical tools.
Related
- A HIP-Compatible Accelerator Backend for Fourier-Bessel Particle-in-Cell Simulations on CPU/DCU Heterogeneous Clusters
- OPALX: A performance-portable framework for particle accelerator simulations
- A particle-in-cell model of beam dynamics in a dielectric wall accelerator
- From Automated Simulation to Autonomous Discovery: A Hierarchical Framework for Agentic Computational Materials Science
- Questioning your brilliance in physics: Differential shifts in fixed mindsets by grade and gender
BibTeX
@misc{wan2026brilliance,
title = {{Brilliance: toward autonomous computational research in particle accelerator physics}},
author = {Jinyu Wan and Yu Zhao and Wei Bao and Ying Li and Hao Zhong and Xiaohan Lu and Yi Jiao},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.15615},
eprint = {2610.15615},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.15615},
}
RIS
TY - UNPB
TI - Brilliance: toward autonomous computational research in particle accelerator physics
AU - Jinyu Wan
AU - Yu Zhao
AU - Wei Bao
AU - Ying Li
AU - Hao Zhong
AU - Xiaohan Lu
AU - Yi Jiao
PY - 2026
DA - 2026/10/28
PB - arXiv
DO - 10.48550/arxiv.2610.15615
UR - https://arxiv.org/abs/2610.15615
AB - Computational research in physics requires continuous interplay between simulation and scientific interpretation, demanding substantial specialist effort. Here we introduce Brilliance, a research agent that combines large language models, reusable physics knowledge and numerical simulation to conduct adaptive investigations in accelerator physics. Given a single user prompt, Brilliance designs a fourth-generation synchrotron light source that meets coupled emittance and nonlinear-acceptance requirements. Further exploration yields a design combining sub-10-pm emittance with a large dynamic aperture under realistic engineering constraints, suggesting a route towards hard-X-ray diffraction-limited operation, although the momentum-acceptance target remains unmet. Brilliance also calibrates a model of an operating light source against measured responses, accurately inferring its lattice and calibration errors. Across these studies, Brilliance autonomously revises its computational strategies in response to numerical evidence. These results demonstrate a route towards delegating sustained scientific investigations to agents equipped with domain expertise and numerical tools.
ER -