mulliganMIT
sim-square-narrow-c01-dagger-sobol-with-cf
Part of the Mulligan dataset collection, this is a training-view variant of the sim-square-narrow task using DAGGER with Sobol sampling and curriculum finetuning, containing 100 episodes of state-only robot control data.
Downloads271
Episodes100
Why This Matters for Physical AI
This dataset contributes to physical AI research by providing a curated, state-only training dataset with explicit provenance tracking and curriculum learning variants that enable systematic evaluation of imitation learning algorithms in robotic manipulation tasks.
Technical Profile
- Modalities
- proprioception
- Environment
- simulation
- Task Types
- manipulation
- Episodes
- 100
- Data Format
- parquet
- Annotation Types
- reward_labels
- License
- MIT
Access
Need custom proprioception data?
Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.
Request a Sample Pack