1 paper across 1 session
This paper mitigates object hallucinations in Large Vision-Language Models by exploring the uncertain visual tokens.