gpt-6VIDEO
GPT-6 Astra learns dexterous manipulation through simulation
The author tested Astra's zero-shot dexterous manipulation and had it learn pen spinning, Rubik's Cube solving, and hammer use through simulation training, with the code open-sourced.
Jianglong Ye
@jianglong_yeArchived 2026.09.18Originally posted 2026.09.18
Jianglong Ye @jianglong_yeOriginally posted 2026.09.18
Evolving dexterity with GPT-6 Astra 🖐️ Been trying Astra recently. Its zero-shot dexterous manipulation is already quite surprising. More interesting is seeing it learn and improve through simulation training, from pen spinning and Rubik's Cube to hammer use. The real goal would be to evolve this dexterity in the real world. https://github.com/jianglongye/dexterous-astra
x.com/jianglong_ye/status/2100617525407211679Archived 2026.09.18 10:09