Search papers, labs, and topics across Lattice.
1
0
3
2
Dataset distillation is more effective when you focus on *what the model can currently learn*, not just visual diversity or average training dynamics.
