2 papers
cs.LG2025
Gradient Alignment in Physics-informed Neural Networks: A Second-Order Optimization Perspective
Sifan Wang, Ananyae Kumar Bhartari, Bowen Li +1
Multi-task learning through composite loss functions is fundamental to modern deep learning, yet optimizing competing objectives remains challenging. We present new theoretical and…
cs.LG2025
Mollifier Layers: Enabling Efficient High-Order Derivatives in Inverse PDE Learning
Ananyae Kumar Bhartari, Vinayak Vinayak, Vivek B Shenoy
Parameter estimation in inverse problems involving partial differential equations (PDEs) underpins modeling across scientific disciplines, especially when parameters vary in space…