Fetching the paper…

MRAG-Bench: Vision-Centric Evaluation for Retrieval-Augmented Multimodal Models · Around