Vchitect/Uni-MMMU-Eval
收藏Hugging Face2025-10-22 更新2025-10-18 收录
下载链接：
https://hf-mirror.com/datasets/Vchitect/Uni-MMMU-Eval
下载链接
链接失效反馈官方服务：
资源简介：
Uni-MMMU是一个大规模多学科多模态统一基准数据集，旨在评估统一模型在视觉理解和生成方面的双向协同作用。它包含八个以推理为中心的领域，如科学、编程、数学和谜题，要求模型利用概念理解来指导精确的视觉合成，或者将生成作为分析推理的认知支架。
Uni-MMMU is a massive multi-discipline multimodal unified benchmark designed to evaluate the bidirectional synergy between visual understanding and generation in unified models. It encompasses eight reasoning-centric domains such as science, coding, mathematics, and puzzles, requiring models to leverage conceptual understanding to guide precise visual synthesis or use generation as a cognitive scaffold for analytical reasoning.
提供机构：
Vchitect
