Robot Learning Glossary
Key concepts in imitation learning, reinforcement learning, and embodied AI — with links to datasets, models, and our data services.
Concepts
Terms & Definitions
Core
Imitation Learning
Learning from demonstrations. Behavior cloning, DAgger, inverse RL. How robots learn from human teleoperation data.
Read more → CoreReinforcement Learning
Learning from trial and error. Reward signals, policy optimization. RL for robotics and sim-to-real.
Read more → ModelsVLA & VLM
Vision-Language-Action and Vision-Language Models. How VLMs enable language-conditioned robot control.
Read more → TransferSim-to-Real Transfer
Training in simulation, deploying in the real world. Domain randomization, reality gap, real-world data.
Read more → DataTeleoperation
Human-in-the-loop control for data collection. Bimanual, mobile, haptic. ALOHA, Mobile ALOHA.
Read more → CorePolicy Learning
Mapping observations to actions. ACT, Diffusion Policy, behavior cloning. Visuomotor policies.
Read more →