This code contains the PINN, PINN-SC, and PINN-Proj models used in the above paper. The code from the workshop paper "Guaranteeing Integrals with Projection in Physics-Informed Neural Networks" can be found in the workshop_code folder.
Generate data using files in data/ folder, use run.sh to train and evaluate PINNs.
Feel free to open an issue if you find a bug or have a question!
@article{baez2025guaranteeing, title={Guaranteeing Conservation of Integrals with Projection in Physics-Informed Neural Networks}, author={Baez, Anthony and Zhang, Wang and Ma, Ziwen and Nguyen, Lam and Das, Subhro and Daniel, Luca}, journal={arXiv preprint arXiv:2511.09048}, year={2025} }