Search papers, labs, and topics across Lattice.
1
0
3
5
Online reinforcement fine-tuning can measurably improve the safety and efficiency of diffusion-based multi-agent driving planners, even when starting from strong pre-trained models.
