Search papers, labs, and topics across Lattice.
1
0
2
Causal priors dramatically improve the robustness of automated feature engineering, reducing performance drops by 4x under distribution shift compared to non-causal methods.
