Res-Bench: Benchmarking the Robustness of Multimodal Large Language Models to Dynamic Resolution Input

Open in new window