KDD 2025 - Can Large Language Models Improve the Adversarial Robustness of Graph Neural Networks?