Search papers, labs, and topics across Lattice.
Affiliation:
2
0
5
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.
