Fetching the paper…

MMLongBench: Benchmarking Long-Context Vision-Language Models Effectively and Thoroughly · Around