SEED-Bench

SEED-Bench Card Benchmark details Benchmark type: SEED-Bench is a large-scale benchmark to evaluate Multimodal Large Language Models (MLLMs). It consists of 19K multiple choice questions with accurate human annotations, which covers 12 evaluation dimensions including the comprehension of both the image and video modality. Benchmark date: SEED-Bench was collected in July 2023. Paper or resources for more information: https://github.com/AILab-CVC/SEED-Bench License:… See the full description on the dataset page: https://huggingface.co/datasets/AILab-CVC/SEED-Bench.

Papers using SEED-Bench (1)

Make Your LVLM KV Cache More Lightweight2026