Skip to content

ConclusionsðŸŠķ

ðŸ”Ū Future Implementations
  • Explore ways to implement adaptive sampling.
  • Train models with learnable frequency \(\omega\).
  • Explicitly condition the network on \(\omega\).
  • Explore richer low-fidelity physics contraints and Hamiltonian structure preservation (e.g., an energy conservation loss term).
  • Explore ways to introduce inductive biases (limitations).