Bērziņš , J.; Kalniņa, E. Robustness of Pre-Trained Language Models Against Adversarial Attacks. MZCJ 2024, 5.