Full Professor, Harvard Medical School, Harvard University
1 paper at NeurIPS 2025
We introduce generative distribution embeddings (GDE), a framework that lifts autoencoders to the space of distributions.