Emerging Applications of Normalizing Flows in Reinforcement Learning
This is Part 2 on Normalizing Flows.

Because of the discussion in part 1, I want to talk about:

  • Requirements on the mapping
    • Map does not need to be bijective
    • Contractions -> useful for high dim cases, e.g. images?
  • Experiment (Mischa)
    • Decoder part is injective I.e. bijective on image
      1. Train encoder as VAE
      2. Train decoder as flow (fine-tune)
  • Convolutional flow networks in more detail
    • How does Glow work? Is it bijective? Same I/O dimensions? (no)
  • Piecewise invertible transformations for flow
  • Dimensionality reduction flows

In this series →

Generative Models
Seminar: Normalizing Flows for Policy … Seminar: An Introduction to Normalizing Flows Seminar: The Levenshtein Transformer Seminar: Part 2: An Introduction to Image … Seminar: An Introduction to Image Synthesis with … Seminar: A Style-Based Generator Architecture for … Seminar: BERT: Pre-training of Deep Bidirectional … Seminar: Principled methods for training GANs