thuml/VisWorld-Eval
收藏Hugging Face2026-03-09 更新2026-02-07 收录
下载链接：
https://hf-mirror.com/datasets/thuml/VisWorld-Eval
下载链接
链接失效反馈官方服务：
资源简介：
VisWorld-Eval套件用于评估具有视觉世界建模的多模态推理能力。它包含七个任务，涵盖合成和现实世界领域，每个任务都设计用于隔离并要求特定的原子世界模型能力。这些任务包括纸张折叠、多跳操作、球体追踪、迷宫、推箱子、立方体三视图投影和现实世界空间推理。数据集是多模态的，涉及推理，并与世界模型相关。
The VisWorld-Eval suite is for assessing multimodal reasoning with visual world modeling. It comprises seven tasks spanning both synthetic and real-world domains, each designed to isolate and demand specific atomic world-model capabilities. These tasks include paper folding, multi-hop manipulation, ball tracking, maze, sokoban, cube 3-view projection, and real-world spatial reasoning. The dataset is multimodal, involves reasoning, and is related to world models.
提供机构：
thuml
