Imitation learning (IL), also called learning from demonstration, trains a policy from examples of an expert performing the task. Osa et al.'s survey "An Algorithmic Perspective on Imitation Learning" (2018) defines it directly: "This process of learning from demonstrations, and the study of algorithms to do so, is called imitation learning." The alternatives are hand-engineering a controller and reinforcement learning, which optimizes a reward signal through trial and error.
IL methods fall into three broad groups. Behavioral cloning treats the problem as supervised learning on recorded observation–action pairs. Interactive methods such as DAgger query the expert on states the learner itself visits. Inverse reinforcement learning infers a reward function from demonstrations and then optimizes it.
Nearly every manipulation policy in this matrix, including ACT, Diffusion Policy, SmolVLA and the fine-tuning stage of π0, is trained by behavioral cloning on demonstrations collected through teleoperation, typically with a leader arm, and stored as a LeRobotDataset. The data requirement can be small: the ACT paper's ALOHA tasks used 50 demonstrations each for most tasks, "around 10-20 minutes of data for each task".
A common misconception is that IL is limited only by the demonstrator's skill. The deeper limit is distribution shift: a cloned policy fails in states its demonstrations never covered, and its own errors lead it there. Pure IL also has no mechanism to improve on the demonstrations. Methods such as HIL-SERL add reinforcement learning with a reward model and report higher success and faster execution than imitation-only baselines (see human-in-the-loop RL).