Mitigating Hallucinations in Large Vision-Language Models by Adaptively Constraining Information Flow

VLM
Re-assign community