Tag
Introduces MultivationBench, a benchmark for evaluating multimodal large language models' sequential motivation reasoning using story-driven visual narratives based on Maslow's hierarchy and Reiss's desires. Results show all tested models struggle with dynamic motivation inference.