Estimating Domain Counterfactuals for Invertible Latent Causal Models without Recovering the Full Causal Structure
It is possible to estimate domain counterfactuals without recovering the full latent causal structure by leveraging the invertibility and sparsity of the causal mechanisms.
