1 paper
Liufan Tan, Jiale Li, Gangshan Jing
Memory-augmented robotic policies are essential in handling memory-dependent tasks. However, existing approaches typically rely on simply extending the observation window, struggli…