2 papers
cs.LG2026
Derived Fields Preserve Fine-Scale Detail in Budgeted Neural Simulators
Wenshuo Wang, Fan Zhang
Fine-scale-faithful neural simulation under fixed storage budgets remains challenging. Many existing methods reduce high-frequency error by improving architectures, training object…
cs.LG2026
Gradient-Informed Temporal Sampling Improves Rollout Accuracy in PDE Surrogate Training
Wenshuo Wang, Fan Zhang
Researchers train neural simulators on uniformly sampled numerical simulation data. But under the same budget, does systematically sampled data provide the most effective informati…