TY  - RPRT
TI  - Benchmarking MLLMs via Cognitive Expected Scene Graph for Safety-Critical Visual Negation Understanding
AU  - Zhiyun Jiang
AU  - Hanyong Wang
AU  - Binbin Liang
AU  - Yu Xie
AU  - Menglong Yang
AU  - Wei Li
PY  - 2026
UR  - https://arxiv.org/abs/2609.19767
ID  - 2609.19767
ER  -