Thus we get the searched architecture and optimized parameters in a single run.

Forpredictivemodels toprovidereliable guidance indecision making processes, they are often required to be accurate and robust to distribution shifts. Shortcut learning-where a model relies on spurious correlations or shortcuts to predict thetargetlabel-undermines therobustnessproperty,leadingtomodelswithpoor out-of-distribution accuracy despite good in-distribution performance.

These algorithms typically recoveracausal graph uptoits Markovequivalence class anditisknowntobefundamentally impossible tolearn causal relationships solely based on observational data.