arxivstat.MLcs.LG2026-07-05
Optimal Mixture-of-Experts Model Averaging for Conditional Generative Models
Shijin Gong, Baihua He, Xinyu Zhang
Conditional generative models have emerged as powerful tools for sampling from target conditional distributions, driving substantial advances across a wide range of scientific and applied domains. As these models proliferate, practitioners often face multiple plausible generators…