Search papers, labs, and topics across Lattice.
University College London
1
0
3
Random splits in urban representation learning benchmarks inflate performance and alter model rankings, making cross-city generalization claims unreliable.
