Linbo Xiong

Action-Adaptive Continual Learning: Enabling Policy Generalization under Dynamic Action Space

ARC-RL: Self-Evolution Continual Reinforcement Learning via Action Representation Space