Return to Issue Details Efficiently Training Physics-Informed Neural Networks via Anomaly-Aware Optimization Download Download PDF