Skip to content

Generative Diffusion Prior Distillation (GDPD)

Home / Concepts / Generative Diffusion Prior Distillation (GDPD)

Generative Diffusion Prior Distillation (GDPD)

Auto-generated stub. Edit this file to add more details.

A knowledge distillation framework that uses a diffusion-based generative prior to guide partial-context student representations toward full-context teacher distributions.

Why It Matters

Introduces a novel paradigm for knowledge distillation in time-series by modeling partial context as a degraded observation of full-context features using diffusion priors.

Metadata & Links