mulligan2026mit
real-square-d2-c05-dagger-mulligan
A robotics manipulation dataset from the Mulligan project containing 50 episodes of a real square manipulation task with multi-camera observations. Part of a larger benchmark for evaluating robotic policies across different model rounds and training variants.
Downloads75
Episodes50
Why This Matters for Physical AI
This dataset contributes to the Mulligan benchmark for evaluating robotic manipulation policies in real-world settings, supporting research on policy learning and generalization across different training methodologies.
Technical Profile
- Modalities
- rgb
- Environment
- lab
- Task Types
- manipulation
- Episodes
- 50
- Data Format
- parquet
- Annotation Types
- reward_labels
- License
- mit
Access
Need custom rgb data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack