Bias and Generalization in Deep Generative Models: An Empirical Study

Abstract

In high dimensional settings, density estimation algorithms rely crucially on their inductive bias. Despite recent empirical success, the inductive bias of deep generative models is not well understood. In this paper we propose a framework to systematically investigate bias and generalization in deep generative models of images. Inspired by experimental methods from cognitive psychology, we probe each learning algorithm with carefully designed training datasets to characterize when and how existing models generate novel attributes and their combinations. We identify similarities to human psychology and verify that these patterns are consistent across commonly used models and architectures.

Cited by in corpus (13)

• Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting
• A Systematic Survey of Regularization and Normalization in GANs
• A Framework for Demonstrating Practical Quantum Advantage: Racing Quantum against Classical Generative Models
• On Disentangled Representations Learned From Correlated Data
• Understanding the Limitations of Variational Mutual Information Estimators
• Deep Generative Models for Detector Signature Simulation: A Taxonomic Review
• On the Sample Complexity of Quantum Boltzmann Machine Learning
• Can neural networks do arithmetic? A survey on the elementary numerical skills of state-of-the-art deep learning models
• Improving the Fairness of Deep Generative Models without Retraining
• Imperfect ImaGANation: Implications of GANs Exacerbating Biases on Facial Data Augmentation and Snapchat Selfie Lenses
• Variational Disentanglement for Domain Generalization
• Inclusive GAN: Improving Data and Minority Coverage in Generative Models
• Rate-Regularization and Generalization in VAEs