Amap releases λ0, a whole-body foundation model trained on human actions
高德发布首个基于开放世界行为的全身基座模型λ0，推动机器人从“局部学习”迈向“全身动作对齐”
- Who
- Gaode
- What
- released the λ0 whole-body foundation model based on open-world behavior
- When
- Quote
近日，阿里巴巴集团旗下高德正式发布首个基于开放世界行为的全身基座模型λ0。
What happened
Amap announced λ0, a whole-body foundation model that learns coordinated movement, posture, and two-handed manipulation from human action data. According to the Amap-provided report republished by QbitAI, its wearable collection pipeline improves data collection efficiency by approximately 3.4x, while real-robot task success improves by 22.5% over the strongest comparison baseline. The team built the 500-hour HumanVerse-500 dataset and reports zero-shot transfer across 7 real-robot tasks; hardware configuration code and sample data are available on GitHub and Hugging Face.
Read next
See the full storyAmap Releases λ0 Whole-Body Model Trained on Human ActionsThe story so far
Amap released λ0, a whole-body foundation model trained on human action data to learn coordinated movement, posture, and two-handed manipulation, and reported zero-shot transfer across 7 real-robot tasks. In an Amap-provided report republished by QbitAI, the team said its wearable data collection pipeline improved collection efficiency by approximately 3.4x and that λ0 improved real-robot task success by 22.5% over the strongest comparison baseline. The team built the 500-hour HumanVerse-500 dataset; hardware configuration code and sample data are available on GitHub and Hugging Face.
Keep up with AI, one email a day
Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue · 2026-10-24