Asmov-1 Robot Policy Reaches 85% Training with 0.007% Error Rate

Andrew Carr announced that the Asmov-1 robot policy is 85% trained on Cartwheel's cross-embodiment control and scaling laws datasets, achieving an error rate of just 0.007% across all evaluation tracks. The policy is set to be open-sourced next week.

2026-11-01 ~ 2026-11-02 · 2 related posts

• Asmov-1 robot policy 85% trained with 0.007% error rate, open source next week — andrew_n_carr · 2026-11-01
• Asmov-1 robot policy 85% trained with 0.007% eval error, open source next week — ericjang11 · 2026-11-02