Graph Attention Network-Based Surrogate Model for Acceleration and Generalization of Injection Molding Simulations (Jan, 10.1007/s12541-025-01345-z, 2026)

초록

This article has been published containing errors related to the incomplete implementation of revisions during the publication process, which resulted in certain table contents being improperly presented and parts of the main text being omitted. The affected items are Table 3 and the sentence describing Equation (7). The original publication has been corrected. The detailed corrections to be reflected in the article are provided below. [Previous version] Table 3 Tuning range of hyperparameters and results Epoch 300, 350, 400, 450 Layer [3, 4, 5, 6] Head number [3, 4, 5, 6] 350 4 4 [Corrected version] Table 3 Tuning range of hyperparameters and results Epoch {300, 350, 400, 450} Layer {3, 4, 5, 6} Head number {2, 4, 8} 350 4 4 [Previous version] To reduce the sensitivity to values close to zero, a small constant of was used as. [Corrected version] To reduce the sensitivity to values close to zero, a small constant of 0.1 was used as ε. © The Author(s), under exclusive licence to Korean Society for Precision Engineering 2026.