TY - RPRT TI - Solving Robotics Tasks with Prior Demonstration via Exploration-Efficient Deep Reinforcement Learning AU - Chengyandan Shen AU - Christoffer Sloth PY - 2026 UR - https://arxiv.org/abs/2509.04069 ID - 2509.04069 ER -