Latent RL Prior for dexterous manipulation
A two-stage framework reuses control knowledge to learn visual policies from limited demonstrations.
Latent RL Prior is a two-stage framework. It reuses control knowledge for data-efficient visual policy learning.
During privileged pretraining, the system distills large-scale reinforcement learning (RL) rollouts. It uses a flow-matching action decoder conditioned on proprioception and latent task context.
Proprioception is used for conditioning the decoder. A latent task context is also used for conditioning.
During visual adaptation, the decoder is frozen. A visual interface maps images to the latent context using action supervision alone.
This method outperforms full-parameter fine-tuning of and GR00T N1.7 in the low-data regime. In door lifting, it uses fewer demonstrations. In bimanual assembly, it uses fewer demonstrations.
The decoder can be shared across arms. It learns coordinated visual contexts for bimanual manipulation.
Key facts
- Latent RL Prior outperforms and GR00T N1.7 in low-data regimes.
- The method uses fewer demonstrations for door lifting.
- The method uses fewer demonstrations for bimanual assembly.