Y. Zhu, “MoHoBench: Assessing Honesty of Multimodal Large Language Models via Unanswerable Visual Questions”, AAAI, vol. 40, no. 34, pp. 29205–29213, Mar. 2026.