Skip to content

论文HF Daily Papers · arXiv 2611.264443 天前

Mind2Act World基准揭示具身智能推理与执行组合能力断层 ​

本文由 AI 根据公开内容整理,可能有误,请以原文为准:Mind2Act World: A Diagnostic Benchmark for Embodied Mind, Physical Action, and Mind-Action Composition in Robotic Manipulation(HF Daily Papers · arXiv 2611.26444)

核心内容 ​

针对现有评测无法区分推理、执行及二者组合失效的问题,研究发布Mind2Act World诊断基准。该基准覆盖5大真实应用领域共30个仿真操作任务,设3级难度与Clean/Random双场景,生成180种配置。通过Mind(历史决策)、Act(精准执行)、Mind2Act(闭环组合)三视图解耦评估,并配套诊断工具将失败归因于特定能力缺失或组合障碍,为多模态智能体、VLA及WAM模型提供细粒度性能画像。

评测结果显示,单视图能力无法迁移至组合任务:最强智能体在Mind和Act视图成功率分别为52.7%和39.0%,但在Mind2Act视图中骤降至12.0%。诊断表明,强推理模型虽能保持任务状态却难以维持持续接触,而端到端策略虽可执行熟练动作却丢失任务状态。这证实通过环境反馈实现推理与执行的动态组合,仍是当前具身智能系统亟待突破的核心工程难题。

关键看点 ​

  • 基准规模:涵盖5大领域30个仿真任务,含3级难度与2种场景设置,总计180种评测配置
  • 性能断层:最强智能体单项推理与执行成功率超39%,但推理-执行组合任务成功率仅12.0%
  • 失效归因:诊断工具可将失败 episode 精确归因为 Mind、Act 或二者组合缺陷,并生成系统能力卡片
  • 模型差异:强推理模型保持状态但接触不稳,端到端策略动作熟练但丢失任务状态记忆

工程与产业观察 ​

该基准量化揭示了“会想”与“会做”在闭环交互中的巨大鸿沟。对于算法工程师,单纯提升VLA或WAM的单模态指标已触及天花板,需转向基于环境反馈的组合机制设计;对于硬件与仿真开发者,应强化支持长时序接触与状态保持的高保真测试环境,以支撑组合能力的迭代验证。

英文摘要 (Abstract) ​

Embodied intelligence aims to let robots perform complex physical tasks with human-like adaptability and reliability, which requires three complementary capabilities: history-based cognitive reasoning, precise and timely physical execution, and the composition of reasoning and execution through environmental feedback. Yet existing benchmarks either evaluate reasoning and execution within the same tasks or assess them separately without their composition, so they reveal little about whether a failure comes from reasoning, execution, or their composition. We introduce Mind2Act World, a diagnostic benchmark of 30 simulated manipulation tasks from five real-world application domains. Each task has three difficulty levels and two scene settings (Clean and Random), yielding 180 configurations. The tasks form three views: Mind tests decisions based on past observations, Act tests precise and timely execution, and Mind2Act combines both in closed-loop tasks where action outcomes guide later decisions. Comparing the three views shows whether a system lacks a capability or fails to compose capabilities it already has. A diagnosis harness further attributes the failed episodes of three representative systems to Mind, Act, or their composition and summarizes each system in a profile card. Our evaluation of multimodal agents, vision-language-action models (VLAs), and world-action models (WAMs) shows that competence in a single view does not carry over to Mind2Act tasks. Even the strongest agent succeeds in 52.7% of Mind and 39.0% of Act episodes but in only 12.0% of Mind2Act episodes. The strongest agent retains task state but struggles with sustained contact, whereas end-to-end policies execute practiced motions but retain little task state. These findings point to composing reasoning and execution through environmental feedback as the main open problem.

Hugging Face Daily Papers 热度:46 upvotes 作者:yuzhi huang, Jie Wu, Fanding Huang, Junchao Lin, Weichen Zhang, ZiYi Xiong, Yang Li, Zhengwei Xie, Zhimin Wang, Zhen Liu, Jiazhen Huang, Yuzheng Zhang, Yuche Liu, Hongyi Yang, Chen Tang, Shixiang Tang…

基于本页技术方向关联的整机参数、核心零部件、实测报告与同赛道选型横评。

外部来源

本页事实对应的公开页面,以原文为准。(点击将离开本站)

  1. HF Daily Papers · arXiv 2611.26444:Mind2Act World: A Diagnostic Benchmark for Embodied Mind, Physical Action, and Mind-Action Composition in Robotic Manipulation ↗访问 2026-11-21