Bare-bones, minified versions of some common (and not-so-common) generative models, for pedagogical purposes.
- β-VAEs: Variational autoencoders and basic rate-distortion theory.
- Diffusion models: Diffusion models, covering likelihood-based and score-matching interpretations.
- Continuous normalizing flows: Continuous-time normalizing flows from e.g., Grathwohl et al 2018.
- Consistency models (WiP annotations): Consistency models from Song et al 2023.
- Flow matching (WiP annotations): From Lipman et al 2022; see also Albergo et al 2023.
- Diffusion distillation (WiP): Progressive (Salimans et al 2022) and consistency (Song et al 2023) distillation.