Physical AI insights

22 articles

Robot Learning, Part 1: Manipulation, From Grasping to Dexterity

Jan 14, 202613 min read

Robot Learning, Part 2: Locomotion, Teaching Robots to Walk and Run

Jan 28, 202613 min read

Robot Learning, Part 3: Imitation Learning, Teaching by Demonstration

Feb 11, 202614 min read

Robot Learning, Part 4: Reinforcement Learning, Skill Through Trial and Error

Feb 25, 202614 min read

Robot Learning, Part 5: Sim-to-Real, Crossing the Reality Gap

Mar 11, 202613 min read

Robot Learning, Part 6: Whole-Body Control and the Road Ahead

Mar 25, 202613 min read

VLA Models, Part 1: The Architecture Behind Robot Learning

Feb 20, 202611 min read

VLA Models, Part 2: How Transformers Fuse Vision, Language, and Action

Mar 12, 202613 min read

VLA Models, Part 3: Tokenizing Action, Turning Motion into a Vocabulary

Mar 26, 202614 min read

VLA Models, Part 4: Training, Data, Losses, and the Fine-Tuning Recipe

Apr 9, 202615 min read

VLA Models, Part 5: Inference and the Control Loop, From Tokens to Torque

Apr 23, 202614 min read

VLA Models, Part 6: Evaluation, Benchmarks, and the Open Problems

May 7, 202613 min read

LeRobot, Part 1: The Open Stack for Real-World Robot Learning

Mar 4, 202612 min read

LeRobot, Part 2: Hardware, Calibration, and Teleoperation

Mar 18, 202612 min read

LeRobot, Part 3: Inside the LeRobotDataset Format

Apr 1, 202613 min read

LeRobot, Part 4: Recording, Tools, and Annotation

Apr 15, 202612 min read

LeRobot, Part 5: Imitation Learning with ACT and Diffusion

Apr 29, 202613 min read

LeRobot, Part 6: Fine-Tuning Foundation Models like SmolVLA

May 13, 202613 min read

LeRobot, Part 7: Reinforcement Learning with HIL-SERL

May 27, 202614 min read

LeRobot, Part 8: Simulation and Benchmarks

Jun 10, 202612 min read

LeRobot, Part 9: Real-Time and Asynchronous Inference

Jun 24, 202613 min read

LeRobot, Part 10: Evaluation, Reward Models, and the Frontier

Jul 8, 202612 min read