Research map / Probabilistic and scientific modeling / Differential equations and PDEs
Is $L^2$ Physics Informed Loss Always Suitable for Training Physics Informed Neural Network?
Chuwei Wang, Shanda Li, Di He, Liwei Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace L2 PINN loss with adversarial L-inf training | PINN · 100-dim LQG | L1 relative error: 3.47% -> 0.27% | Improved |
| replace L2 PINN loss with adversarial L-inf training | PINN · 250-dim LQG | L1 relative error: 6.74% -> 0.95% | Improved |
| add curriculum regularization to L2 PINN | PINN · 100/250-dim LQG | L1 relative error: 3.47% -> 3.40% | No clear effect |
| add learning rate annealing to L2 PINN | PINN · 100-dim LQG | L1 relative error: 3.47% -> 11.09% | Hurt |
| use adversarial training instead of directly minimizing L^8 loss (ablation) | L-inf adversarial PINN · 100-dim LQG | L1 relative error: 53.55% -> 0.27% | Improved |
| use adversarial training instead of directly minimizing L^4 loss (ablation) | L-inf adversarial PINN · 100-dim LQG | L1 relative error: 2.42% -> 0.27% | Improved |
| add adversarial training to boundary residual only (ablation) | PINN · 100-dim LQG | L1 relative error: 3.47% -> 2.79% | Improved |
| add adversarial training to both domain and boundary residuals (vs boundary only) (ablation) | L-inf adversarial PINN · 100-dim LQG | L1 relative error: 2.79% -> 0.27% | Improved |
| increase inner-loop iterations K and reduce step size at fixed K*eta (ablation) | L-inf adversarial PINN · 100-dim LQG | L1 relative error | Improved |
Closest papers by meaning
- Efficient Error Certification for Physics-Informed Neural Networks (ICML 2024 · similarity 0.90)
- Learning from Integral Losses in Physics Informed Neural Networks (ICML 2024 · similarity 0.90)
- Challenges in Training PINNs: A Loss Landscape Perspective (ICML 2024 · similarity 0.89)
- PINNsFormer: A Transformer-Based Framework For Physics-Informed Neural Networks (ICLR 2024 · similarity 0.89)
- Dual Cone Gradient Descent for Training Physics-Informed Neural Networks (NeurIPS 2024 · similarity 0.89)
- Sub-Sequential Physics-Informed Learning with State Space Model (ICML 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
