Search papers, labs, and topics across Lattice.
Affiliation:
3
0
5
Value-driven optimistic exploration has historically crumbled under deep network approximations, but diagnosing and repairing failure modes in uncertainty propagation allows pure model-free RL to beat complex model-based exploration baselines.
Action chunking in CRL boosts performance by over 90% in certain environments, revealing that richer temporal representations can significantly enhance learning.
Uncertainty quantification in VLAs can reduce the need for costly expert demonstrations by over 22%, enhancing their adaptability and reliability in real-world applications.
