@misc{indiciaed270ce07e03a,
  title = {Benchmarking MLLMs via Cognitive Expected Scene Graph for Safety-Critical Visual Negation Understanding},
  author = {Zhiyun Jiang and Hanyong Wang and Binbin Liang and Yu Xie and Menglong Yang and Wei Li},
  year = {2026},
  url = {https://arxiv.org/abs/2609.19767},
  note = {Source identifier: 2609.19767}
}