Jiang, W., Liang, K., Rong, X., Zhou, J., Zhong, Z., Wan, G., & Wang, J. (2026). PurMM: Attention-Guided Test-Time Backdoor Purification in Multimodal Large Language Models. Proceedings of the AAAI Conference on Artificial Intelligence, 40(42), 35562–35570. https://doi.org/10.1609/aaai.v40i42.40867