MindTopo基准揭示多模态模型空间推理短板

AI 摘要

微软研究院推出新基准 MindTopo，用于评估多模态大语言模型的拓扑推理能力，涵盖连通性、封闭、顺序、分离和打结等概念。研究发现，当前模型在静态图像识别上表现较好，但在交互式规划任务中显著下降，错误多出现在规划阶段而非感知阶段。该基准旨在为机器人及交互环境中的 AI 系统提供诊断工具，并指出需要显式拓扑状态或保持拓扑的世界模型。

正文节选

At a glance - MindTopo is a new benchmark for testing topological reasoning in AI, evaluating whether multimodal models can understand concepts such as connectivity, enclosure, order, separation, and knots. - The benchmark measures both reasoning and planning, testing not only whether models can recognize topological relationships in static images but also whether they can preserve and manipulate those relationships through a sequence of actions. - Current multimodal models perform much better o

发布时间：2026-08-13 00:00

抓取时间：2026-08-13 00:21

来源机构：Microsoft Research