Explore a comprehensive presentation on discrete diffusion modeling techniques that focus on estimating the ratios of data distribution. Delve into the research paper "Discrete diffusion modeling by estimating the ratios of the data distribution" as Aaron Lou from the Generative Memory Lab provides an in-depth analysis and discussion of its findings. Gain insights into advanced concepts in machine learning and data analysis through this 81-minute talk, which offers a thorough examination of the paper's methodology, results, and implications for the field of discrete diffusion modeling.
Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
Generative Memory Lab via YouTube
Overview
Syllabus
Discrete diffusion modeling by estimating the ratios of the data distribution
Taught by
Generative Memory Lab
Related Courses
-
Blackout Diffusion: Generative Diffusion Models in Discrete-State Spaces
-
Reflected Diffusion Models
-
Deep Generative Models - Diffusion Models for Discrete Data - Lecture 18
-
Understanding Text Generation Through Diffusion Models - From Theory to Implementation
-
Action Matching: Learning Stochastic Dynamics from Samples
-
Generative Modeling by Estimating Gradients of the Data Distribution - Stefano Ermon
