Learning from demonstrations is effective for robotic manipulation, but collecting sufficient task-specific data remains a major bottleneck. Under distribution shift, small errors compound, performance degrades, and expert time is often spent on redundant, low-value corrections instead of the few critical failure cases.

翻译：从示范中学习对机器人操作有效，但收集足够的特定任务数据仍是主要瓶颈。在分布偏移下，小误差会累积，性能下降，专家时间常消耗在冗余、低价值的修正上，而非少量关键失败案例。