Keywords
backdoor defense
adversarial attacks
Abstract
High-stakes learning systems face security challenges from adversarial attacks, poisoned training data, and model-transfer vulnerabilities. CLIP-based models are especially important because they connect visual and textual representations across many downstream tasks. Universal transferable attacks reveal how one surrogate model can generate targeted adversarial effects across systems, while entropy-based poisoned dataset separation offers a defense-oriented approach to CLIP-guided backdoor detection. Sequential recommendation with LLMs adds a personalization dimension, where dual dynamic indexing can improve recommendation performance but also increases dependence on reliable data and model behavior. Biomedical injury evidence and flood relocation research provide high-stakes application contexts in which model failure, uncertainty, or manipulation can have serious consequences. This literature cluster supports the study of secure, adaptive, and context-aware intelligent systems.
References
Xu, B., Dai, X., Tang, D., & Zhang, K. (2025, November). One surrogate to fool them all: Universal, transferable, and targeted adversarial attacks with clip. In Proceedings of the 2025 ACM SIGSAC Conference on Computer and Communications Security (pp. 3087-3101).
Xu, B., Yang, F., Dai, X., Tang, D., & Zhang, K. (2025, October). CLIP-Guided Backdoor Defense through Entropy-Based Poisoned Dataset Separation. In Proceedings of the 33rd ACM International Conference on Multimedia (pp. 7415-7423).
Yin, J., Zeng, Z., Li, M., Yan, H., Li, C., Han, W., ... & Wang, S. (2025, April). Unleash llms potential for sequential recommendation by coordinating dual dynamic index mechanism. In Proceedings of the ACM on Web Conference 2025 (pp. 216-227).
Zhang, Z., Duan, Y., Gu, F., Wei, J., Huo, H., Wang, Y., ... & Gao, Z. (2026). Circulating ATP from hepatic ischemia-reperfusion drives remote cardiac injury via macrophage inflammasome activation. Pharmacological Research, 108306.
Zhou, Y. (2022). Pre-disaster relocation and agent-based model for flood disaster [Doctoral dissertation, University of Wisconsin–Madison].
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., ... & Sutskever, I. (2021). Learning transferable visual models from natural language supervision. In International conference on machine learning (pp. 8748-8763). PMLR.
Goodfellow, I. J., Shlens, J., & Szegedy, C. (2015). Explaining and harnessing adversarial examples. In International Conference on Learning Representations.
Carlini, N., & Wagner, D. (2017). Towards evaluating the robustness of neural networks. In 2017 IEEE Symposium on Security and Privacy (SP) (pp. 39-57). IEEE.
Wang, B., Yao, Y., Shan, S., Li, H., Viswanath, B., Zheng, H., & Zhao, B. Y. (2019). Neural cleanse: Identifying and mitigating backdoor attacks in neural networks. In 2019 IEEE Symposium on Security and Privacy (SP) (pp. 707-723). IEEE.
Kang, W. C., & McAuley, J. (2018, November). Self-attentive sequential recommendation. In 2018 IEEE International Conference on Data Mining (ICDM) (pp. 197-206). IEEE.