1 paper · 1 filter
Sheng Hu, Weiyi Lu, Lingbing Zeng +5
Progress reward modeling is the problem of estimating how a robot's behavior changes task progress over time. Reliable estimation requires distinguishing meaningful state changes f…