A0332
Title: Generative models and approximate Bayesian inference
Authors: Christian Andersson Naesseth - University of Amsterdam (Netherlands) [presenting]
Abstract: Generative models have taken the world by storm. Generative modelling, or generative AI, is the task of constructing an approximation to the data-generating process in the form of a probability distribution. In the context of text, for example, in large language models, the distribution is over words (or tokens), whereas for images, it is an approximate probability distribution over pixel values. The similarities, connections, and potential synergies between generative AI and approximate Bayesian inference are discussed.
