Unifying Generative Models with GFlowNets and Beyond

TL;DR

This paper unifies various deep generative models under the GFlowNet framework, revealing their connections and offering new training and inference strategies to improve generative modeling.

Contribution

It introduces a unifying perspective on deep generative models using GFlowNets and provides practical methods for enhancing generative performance.

Findings

Unified training and inference algorithms for generative models

Experimental validation of improved generative modeling techniques

Insights into the decision-making process in sampling methods

Abstract

There are many frameworks for deep generative modeling, each often presented with their own specific training algorithms and inference methods. Here, we demonstrate the connections between existing deep generative models and the recently introduced GFlowNet framework, a probabilistic inference machine which treats sampling as a decision-making process. This analysis sheds light on their overlapping traits and provides a unifying viewpoint through the lens of learning with Markovian trajectories. Our framework provides a means for unifying training and inference algorithms, and provides a route to shine a unifying light over many generative models. Beyond this, we provide a practical and experimentally verified recipe for improving generative modeling with insights from the GFlowNet perspective.