Notes

The field of machine learning has witnessed a surge in research on generative models, as these models possess the ability to generate novel data similar to the training data. We explore various types of generative models, with a thorough analysis of their respective limitations and potential applications. Additionally, we also investigate several influence methods / instance-based interpretation and explore their efficacy in relevant applications.

Click here to visit our page.


Generative Models

  1. Variational Autoencoders (VAE)
  2. Generative Adversarial Networks (GAN)
  3. Normalizing Flows
  4. Diffusion Models

Influence Methods

  1. TracIn
  2. Influence Functions
  3. Representer Points
  4. (Application: Memorization in Generative Models)