Understanding and Mitigating Hallucination in Large Vision-Language Models via Modular Attribution and Intervention.

Tianyun Yang, Ziniu Li, Juan Cao, Chang Xu

Browse the full ICLR paper archive.