A backpropagation-free training approach for Hamiltonian neural networks via data-driven parameter sampling that claims over 100x CPU speedup and four orders of magnitude better accuracy on chaotic systems like Hénon-Heiles compared to gradient-based methods.
Geometric numerical integration
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 2roles
background 1polarities
background 1representative citing papers
SharpEuler estimates a sharpness profile via finite differences on calibration trajectories, smooths it, and applies a quantile transform to generate adaptive timestep grids that improve Euler sampling quality in flow matching models at fixed budgets.
citing papers explorer
-
Training Hamiltonian neural networks without backpropagation
A backpropagation-free training approach for Hamiltonian neural networks via data-driven parameter sampling that claims over 100x CPU speedup and four orders of magnitude better accuracy on chaotic systems like Hénon-Heiles compared to gradient-based methods.
-
Sharpen Your Flow: Sharpness-Aware Sampling for Flow Matching
SharpEuler estimates a sharpness profile via finite differences on calibration trajectories, smooths it, and applies a quantile transform to generate adaptive timestep grids that improve Euler sampling quality in flow matching models at fixed budgets.