Diffusion and flow processes forget dependencies to define valid copulas then learn to remember them for density estimation and sampling, outperforming prior copula methods on complex datasets.
The diffusion duality
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
Generative perplexity and entropy are shown to be the two additive components of KL divergence to a reference distribution, motivating generative frontiers as a principled evaluation method for diffusion language models.
citing papers explorer
-
Diffusion and Flow-based Copulas: Forgetting and Remembering Dependencies
Diffusion and flow processes forget dependencies to define valid copulas then learn to remember them for density estimation and sampling, outperforming prior copula methods on complex datasets.
-
Generative Frontiers: Why Evaluation Matters for Diffusion Language Models
Generative perplexity and entropy are shown to be the two additive components of KL divergence to a reference distribution, motivating generative frontiers as a principled evaluation method for diffusion language models.