J. Fan and C.-M. Chen, “Efficient Multimodal Large Language Model via Dynamic KV Cache Quantization”, AAAI, vol. 40, no. 25, pp. 20994–21001, Mar. 2026.