Combating Multimodal LLM Hallucination via Bottom-Up Holistic Reasoning

Shengqiong Wu, Hao Fei, Liangming Pan +3

Recent advancements in multimodal large language models (MLLMs) have shown unprecedented capabilities in advancing various vision-language tasks. However, MLLMs face significant ch…