Blackout Diffusion: Generative Diffusion Models in Discrete-State Spaces

Explore a groundbreaking approach to generative diffusion models in discrete-state spaces through this 48-minute conference talk by Yen Ting Lin from the Generative Memory Lab. Delve into the innovative concepts presented in the paper "Blackout Diffusion: Generative Diffusion Models in Discrete-State Spaces," which introduces a novel framework for applying diffusion models to discrete data. Gain insights into the potential applications and implications of this research for various fields, including natural language processing and image generation.