Appendix · Embodiment

Back to core

Appendix: Index · Prev: Reasoning & planning · Next: Autonomy

Overview#copy

Embodiment connects perception and action: policies must map high-dimensional sensory inputs to control, often with safety constraints and limited data.

1) Vision-language-action policies#copy

Key papers#copy

• RT-1: Robotics Transformer for Real-World Control at Scale (Brohan et al., 2022)
• RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control (Brohan et al., 2023)
• PaLM-E: An Embodied Multimodal Language Model (Driess et al., 2023)

Code repositories / projects#copy

• google-research/robotics_transformer

2) Imitation learning and diffusion policies#copy

Key papers#copy

• Diffusion Policy (Chi et al., 2023)

Code repositories / projects#copy

• real-stanford/diffusion_policy

3) Sim-to-real and benchmarking#copy

Key papers#copy

• D4RL: Datasets for Deep Data-Driven Reinforcement Learning (Fu et al., 2020)

Code repositories / projects#copy

• Farama-Foundation/D4RL