Yang Xiao. "Adaptive trust evaluation and representation-based robust aggregation against poisoning attacks in federated learning." Scientific Reports (2026). https://doi.org/10.1038/s41598-026-50985-2