Modern interfaces rely on animation to communicate state, feedback, & intent, but can #AI understand it? This #ACL2026 paper introduces AniMINT, the first benchmark for evaluating Vision Language Models on UI animations: 🔗aclanthology.org/2026.findin…