Photonics-GCCE: group collaborative-competitive evolution multi-agent framework for universal and autonomous optical design
Abstract
Large language model (LLM)-empowered photonic agents connect natural-language intents to executable solvers, showing significant advantages over conventional optical design approaches. However, current multi-agent frameworks operate within a collaborative paradigm without extrinsic selective pressure, which could inherit shared blind spots, converge prematurely, and fail to accumulate transferable experience for intricate tasks. Here, we introduce a group photonics collaboration-compete evolution (GCCE) framework and its LLM instantiation, termed Photonics-GCCE. Two independent agent groups pursue the same design target and undergo structured competitive evaluation across refractive-index fidelity, fabrication sensitivity, algorithmic adequacy, and physical consistency. Each group comprises a leader and three specialist agents dedicated to materials, optimization, and code validation. Agents refine their skills through competitive evaluation across design rounds. Benchmarking across six device categories against single-agent and multi-agent baselines shows that Photonics-GCCE elevates composite scores into the high 90s, improves fabrication robustness by 15 to 17 points, and reduces solver iterations to roughly 40 rounds. A representative quasi?BIC demonstration achieves a practically fabricable design with a quality factor of 13120. Our results demonstrate Photonics-GCCE as a general-purpose and closed-loop framework for autonomous optical design, capable of producing high-performance, fabrication-ready devices across diverse nanophotonic tasks.
Related
- Collaborative AI agents for autonomous experimental research with solid-state quantum spins
- Toward a Thermodynamic Framework for Dissipative Solitons: From Photonics to Turbulence and Bose--Einstein Condensate Analogies
- From Automated Simulation to Autonomous Discovery: A Hierarchical Framework for Agentic Computational Materials Science
- CFD Forge: A Physics-Constrained Agentic Framework for Autonomous CFD
- Thermal Point-Source Reconstruction for Integrated Photonics: Crosstalk Limits and Thermorefractive-Noise Correlations
BibTeX
@misc{xu2026photonicsgcce,
title = {{Photonics-GCCE: group collaborative-competitive evolution multi-agent framework for universal and autonomous optical design}},
author = {Weijie Xu and Ming Wang and Ruicheng Ma and Zeyong Wei and Di Zhang and Jian Chen and Zining Wang and Sijun Hu and Linyuan Dou and Haoyu Li and Haigang Liang and Yijie Luo and Shuqiao Li and Qinghua Song and Huihui Zhu and Hongfei Jiao and Chao Liu and Ali Momeni and Zhanshan Wang and Yuzhi Shi and Romain Fleury and Cheng-Wei Qiu and Xinbin Cheng},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.28045},
eprint = {2609.28045},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.28045},
}
RIS
TY - UNPB
TI - Photonics-GCCE: group collaborative-competitive evolution multi-agent framework for universal and autonomous optical design
AU - Weijie Xu
AU - Ming Wang
AU - Ruicheng Ma
AU - Zeyong Wei
AU - Di Zhang
AU - Jian Chen
AU - Zining Wang
AU - Sijun Hu
AU - Linyuan Dou
AU - Haoyu Li
AU - Haigang Liang
AU - Yijie Luo
AU - Shuqiao Li
AU - Qinghua Song
AU - Huihui Zhu
AU - Hongfei Jiao
AU - Chao Liu
AU - Ali Momeni
AU - Zhanshan Wang
AU - Yuzhi Shi
AU - Romain Fleury
AU - Cheng-Wei Qiu
AU - Xinbin Cheng
PY - 2026
DA - 2026/09/23
PB - arXiv
DO - 10.48550/arxiv.2609.28045
UR - https://arxiv.org/abs/2609.28045
AB - Large language model (LLM)-empowered photonic agents connect natural-language intents to executable solvers, showing significant advantages over conventional optical design approaches. However, current multi-agent frameworks operate within a collaborative paradigm without extrinsic selective pressure, which could inherit shared blind spots, converge prematurely, and fail to accumulate transferable experience for intricate tasks. Here, we introduce a group photonics collaboration-compete evolution (GCCE) framework and its LLM instantiation, termed Photonics-GCCE. Two independent agent groups pursue the same design target and undergo structured competitive evaluation across refractive-index fidelity, fabrication sensitivity, algorithmic adequacy, and physical consistency. Each group comprises a leader and three specialist agents dedicated to materials, optimization, and code validation. Agents refine their skills through competitive evaluation across design rounds. Benchmarking across six device categories against single-agent and multi-agent baselines shows that Photonics-GCCE elevates composite scores into the high 90s, improves fabrication robustness by 15 to 17 points, and reduces solver iterations to roughly 40 rounds. A representative quasi?BIC demonstration achieves a practically fabricable design with a quality factor of 13120. Our results demonstrate Photonics-GCCE as a general-purpose and closed-loop framework for autonomous optical design, capable of producing high-performance, fabrication-ready devices across diverse nanophotonic tasks.
ER -