S. Bondarovets and T. Okhrimenko, “Method for enhancing the robustness of large language models against adversarial attacks”, Bezpeka ìnf., vol. 32, no. 2, pp. 60–71, Aug. 2026.