junwon-seo2026mit
LUCID: Modeling Latent Disturbances for Robust Decision-Making in World Models
Dataset for training and evaluating world models with latent disturbance modeling for robust decision-making in robotic manipulation tasks. Contains offline episodes for block pouring and real robot episodes for egg serving demonstrations.
Downloads19
Episodes2821
Why This Matters for Physical AI
This dataset enables research into robust world models that can handle latent disturbances and model uncertainty, which is critical for deploying manipulation robots in real-world environments where perfect observations are unavailable.
Technical Profile
- Environment
- labsimulation
- Task Types
- manipulationpouringpick_and_place
- Episodes
- 2821
- License
- mit
Access
Need custom physical AI data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack