1 paper
Shi Jin, Yuntian Wang, Yuhui Duan +16
Real-robot evaluation is essential for understanding whether learned manipulation policies can operate reliably outside curated demonstrations. This need is particularly pressing f…