Search papers, labs, and topics across Lattice.
1
0
2
3
RynnValue shows that using temporal distance as a supervision target can outperform traditional preference-based methods in robotic learning, achieving higher accuracy and broader generalization.
