外观
论文arXiv 2609.3853729 天前
系统探索 GPT-6 Astra 作为具身策略的能力
原文链接：Systematically Exploring the Capabilities of GPT-6 Astra as Embodied Policies（arXiv 2609.38537 · 发布日期：2026-09-29）
Systematically Exploring the Capabilities of GPT-6 Astra as Embodied Policies
为评估GPT-6 Astra作为通用具身策略，论文在六个领域测试直接控制、与学习策略协作及反馈适应。
核心要点
- Astra可生成数值机器人动作，需评估其是否能作为通用具身策略。
- 研究考察直接控制、与学习策略协作，以及基于反馈的适应能力。
- 混合控制在夹爪、灵巧和移动操作中成功率为48%、50%和38.7%。
- 导航成功率较高且部分人形任务超越基线，但运动控制可靠性和推理延迟仍受限。
验证范围与局限：验证覆盖六个领域的若干评测子集和任务，包括RoboDojo、DexJoCo、RoboCasa365、RxR、HM3D、HumanoidBench及单个障碍路线。
英文摘要 (Abstract)
GPT-6 Astra exhibits a remarkable ability to generate numerical robot actions, extending its role beyond high-level planning. To assess Astra's capabilities as general-purpose embodied policies, we conduct comprehensive evaluations across six domains, examining direct control, cooperation with learned policies, and feedback-driven adaptation. In gripper manipulation, Astra can correct task targets and prepare contact conditions for subsequent policy execution; hybrid control with π0.5 achieves 48% success on the evaluated RoboDojo subset. In dexterous manipulation, hybrid control achieves 50% success in ten experience-guided DexJoCo trials, while direct in-hand control struggles to coordinate finger contacts. In mobile manipulation, hybrid control reaches 38.7% success on the evaluated RoboCasa365. In navigation, Astra leads our local comparisons, reaching 92% success on RxR instruction following and 82% on HM3D object search, although search incurs substantial detours. In locomotion, dense motion-reference generation remains unreliable: none of five sequential attempts on a single obstacle course reaches the goal, despite improvements in stability and forward progress. In humanoid loco-manipulation, Astra exceeds baseline methods on 13 of 30 HumanoidBench tasks with pretrained whole-body controllers. These findings reveal a gap between useful task decisions and reliable physical control. Inference latency further constrains practical control: across 50 RoboDojo instances per condition, policy-assisted and direct control consume 624.8 million and 1.132 billion tokens. A 30-second locomotion run requires 250 model calls averaging 39.86 seconds each, with physics paused during inference.
作者：Galbot Team, Xuchuan Chen, Xiaoqian Cheng, Yu Deng, Lihe Ding, Shaocong Dong, Xiangjun Gao, Haozhe Jia, Zekai Li, Zhoujian Li
来源
本页事实对应的公开页面，以原文为准。