1 paper
Xiaofeng Zhang, Yuanchao Zhu, Chaochen Gu +8
Recent studies have examined attention dynamics in large vision-language models (LVLMs) to detect hallucinations. However, existing approaches remain limited in reliably distinguis…