作者位次 #2/6ICML 2026Poster· Deep Learning / Foundation Models
Finding the Correct Visual Evidence Without Forgetting: Mitigating Hallucination in LVLMs via Inter-Layer Visual Attention Discrepancy
Yutong Xie · Zhenglin Hua · Ran Wang · Wing W. Y. Ng · Xizhao Wang · Yuheng Jia