GPT-6Astra Embodied Intelligence Breakthrough and Physical AI Data Flywheel Initiation
开源证券:GPT-6Astra具身智能突破 物理AI数据飞轮开启
A research paper evaluates GPT-6Astra's performance in embodied intelligence tasks, achieving high success rates on benchmarks like LIBERO-Long++, SIMPLE, and DexJoCo. The study also explores the implications of these results for physical AI self-evolution and the growing data demands in the field.
The study evaluates GPT-6Astra's embodied intelligence capabilities across multiple benchmarks, showing significant improvements over previous models. It highlights the potential for physical AI self-evolution through data-driven training and simulation.
Source: Chinese robotics — Dataset and collection discovery · Read original article ↗
English translation unavailable. Choose the original article to read it.
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
Reported success rate
100%
View original evidence
》对GPT-6Astra的具身能力进行测试,通过特定的条件(In-contextlearning+最高10倍步数预算)测试,发现在长程机器人操作基准LIBERO-Long++评测上成功率达100%(此前SOTA82%),操作鲁棒性和泛化能力的LIBERO-Pro上成功率99.3%(此前SOTA50%),人形机器人基准SIMPLE成功率达98.8%(此前SOTA92.5%),灵巧操作DexJoCo成功率87.3%(此前SOTA40.3%)。
Open source S4Reported success rate
99.3%
View original evidence
》对GPT-6Astra的具身能力进行测试,通过特定的条件(In-contextlearning+最高10倍步数预算)测试,发现在长程机器人操作基准LIBERO-Long++评测上成功率达100%(此前SOTA82%),操作鲁棒性和泛化能力的LIBERO-Pro上成功率99.3%(此前SOTA50%),人形机器人基准SIMPLE成功率达98.8%(此前SOTA92.5%),灵巧操作DexJoCo成功率87.3%(此前SOTA40.3%)。
Open source S4Reported success rate
98.8%
View original evidence
》对GPT-6Astra的具身能力进行测试,通过特定的条件(In-contextlearning+最高10倍步数预算)测试,发现在长程机器人操作基准LIBERO-Long++评测上成功率达100%(此前SOTA82%),操作鲁棒性和泛化能力的LIBERO-Pro上成功率99.3%(此前SOTA50%),人形机器人基准SIMPLE成功率达98.8%(此前SOTA92.5%),灵巧操作DexJoCo成功率87.3%(此前SOTA40.3%)。
Open source S4Reported success rate
87.3%
View original evidence
》对GPT-6Astra的具身能力进行测试,通过特定的条件(In-contextlearning+最高10倍步数预算)测试,发现在长程机器人操作基准LIBERO-Long++评测上成功率达100%(此前SOTA82%),操作鲁棒性和泛化能力的LIBERO-Pro上成功率99.3%(此前SOTA50%),人形机器人基准SIMPLE成功率达98.8%(此前SOTA92.5%),灵巧操作DexJoCo成功率87.3%(此前SOTA40.3%)。
Open source S4Reported success rate
95%
Reported zero-shot result; the excerpt does not specify an evaluation protocol.
View original evidence
破壳机器人创始人许华哲表示,GPT-6Astra在机器人控制任务中成功率达95%,前代模型Fable5.1只有40%,在RoboLab语义理解任务中,Astra直接控制的零样本成功率为98%。
Open source S5Reported success rate
98%
Reported zero-shot result; the excerpt does not specify an evaluation protocol.
View original evidence
破壳机器人创始人许华哲表示,GPT-6Astra在机器人控制任务中成功率达95%,前代模型Fable5.1只有40%,在RoboLab语义理解任务中,Astra直接控制的零样本成功率为98%。
Open source S5
- paper: not_reported
What remains unknown
Not established in the collected evidence: Environment, Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
》对GPT-6Astra的具身能力进行测试,通过特定的条件(In-contextlearning+最高10倍步数预算)测试,发现在长程机器人操作基准LIBERO-Long++评测上成功率达100%(此前SOTA82%),操作鲁棒性和泛化能力的LIBERO-Pro上成功率99.3%(此前SOTA50%),人形机器人基准SIMPLE成功率达98.8%(此前SOTA92.5%),灵巧操作DexJoCo成功率87.3%(此前SOTA40.3%)。
Open source S4
破壳机器人创始人许华哲表示,GPT-6Astra在机器人控制任务中成功率达95%,前代模型Fable5.1只有40%,在RoboLab语义理解任务中,Astra直接控制的零样本成功率为98%。
Open source S5
Source:Chinese robotics — Dataset and collection discovery · news.qq.com