Conditional Generative Models are Sufficient to Sample from Any Causal Effect Estimand Md Musfiqur Rahman Purdue University Matt Jordan University of Texas at Austin Murat Kocaoglu Purdue University

Nov-19-2025, 20:53:11 GMT–Neural Information Processing Systems 

While sound and complete algorithms exist to compute causal effects, many of them assume access to conditional likelihoods, which is difficult to estimate for high-dimensional (particularly image) data.