ResearchBenchmarksGlobal error metrics prediction on Case 1 (full-state train)Follow2.4625Mean RMSEPhyCNN2.180364.0848055.989257.893695Nov 26, 2025Evaluation ResultsMethodMethodLinksMean RMSEMedian RMSEMean Rel-L2Median Rel-L2PhyCNNtraining_mode=full-sta...training_mode=full-state training2025.112.46251.681447.441.77PhyULSTMtraining_mode=full-sta...training_mode=full-state training2025.119.5165.685619.2815.6