Rong X, Huang W, Jiang W, Li Y, Wang W, Ye M. Probing Semantic Insensitivity for Inference-Time Backdoor Defense in Multimodal Large Language Model. AAAI [Internet]. 2026 Mar. 14 [cited 2026 May 27];40(42):35775-83. Available from: https://ojs.aaai.org/index.php/AAAI/article/view/40891