July 31, 2025

Quantum Physics-Informed Neural Networks for Maxwell’s Equations: Circuit Design, “Black Hole” Barren Plateaus Mitigation, and GPU Acceleration

View Full Paper

Ask AI

Bookmark

Authors

ZCZu-Cheng ChenHunan Normal University GSGal G. ShavinerTechnion – Israel Institute of Technology HCHemanth ChandravamsiTechnion – Israel Institute of Technology

Discussion

Member takes

Implication

Quantum physics-informed neural networks enhance accuracy of Maxwell’s equations solutions, suggesting effective training methods for electromagnetic applications.

Key Points

The proposed quantum physics-informed neural network improves solving Maxwell’s equations with reduced parameters.
An energy conservation term in the loss function stabilizes training, enhancing solution fidelity.
The method showed higher accuracy in 2D wave propagation problems in both vacuum and dielectric media.
Incorporating advanced techniques mitigates barren plateaus in the training landscape, optimizing performance.

Cite This Study

Chen et al. (2025) studied this question.

synapsesocial.com/papers/689a0c5fe6551bb0af8cf747 https://doi.org/10.21203/rs.3.rs-7010088/v1

View Full Paper

Ask AI

Bookmark

Implication

Quantum physics-informed neural networks enhance accuracy of Maxwell’s equations solutions, suggesting effective training methods for electromagnetic applications.

Quantum Physics-Informed Neural Networks for Maxwell’s Equations: Circuit Design, “Black Hole” Barren Plateaus Mitigation, and GPU Acceleration

Discussion

Member takes

Implication

Key Points

Cite This Study

Quantum Physics-Informed Neural Networks for Maxwell’s Equations: Circuit Design, “Black Hole” Barren Plateaus Mitigation, and GPU Acceleration

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider