Bias and Generalization in Deep Generative Models: An Empirical Study
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.
Authors

Are you an author of this paper? Check the Twitter handle we have for you is correct.

Shengjia Zhao (edit)
Hongyu Ren (add twitter)
Arianna Yuan (add twitter)
Jiaming Song (edit)
Noah Goodman (edit)
Stefano Ermon (edit)
Ask The Authors

Ask the authors of this paper a question or leave a comment.

Read it. Rate it.
#1. Which part of the paper did you read?

#2. The paper contains new data or analyses that is openly accessible?
#3. The conclusion is supported by the data and analyses?
#4. The conclusion is of scientific interest?
#5. The result is likely to lead to future research?

Github
Stargazers:
3
Forks:
1
Open Issues:
0
Network:
1
Subscribers:
2
Language:
Jupyter Notebook
Youtube
Link:
None (add)
Views:
0
Likes:
0
Dislikes:
0
Favorites:
0
Comments:
0
Other
Sample Sizes (N=):
Inserted:
Words Total:
Words Unique:
Source:
Abstract:
None
11/08/18 06:04PM
8,814
2,073
Tweets
sheath_38: RT @arxiv_org: Bias and Generalization in Deep Generative Models: An Empirical Study. https://t.co/MQDVUd5ZYl https://t.co/8yzgRRgRBW
hereticreader: [1811.03259] Bias and Generalization in Deep Generative Models: An Empirical Study - https://t.co/9D1COPFWPY https://t.co/9YfBw583PN
arxiv_org: Bias and Generalization in Deep Generative Models: An Empirical Study. https://t.co/MQDVUd5ZYl https://t.co/8yzgRRgRBW
StatsPapers: Bias and Generalization in Deep Generative Models: An Empirical Study. https://t.co/WwPrHP9svM
PerthMLGroup: RT @Miles_Brundage: "Bias and Generalization in Deep Generative Models: An Empirical Study," Zhao and Ren et al.: https://t.co/26CwrN9Xu7
heghbalz: RT @Miles_Brundage: "Bias and Generalization in Deep Generative Models: An Empirical Study," Zhao and Ren et al.: https://t.co/26CwrN9Xu7
AssistedEvolve: RT @Miles_Brundage: "Bias and Generalization in Deep Generative Models: An Empirical Study," Zhao and Ren et al.: https://t.co/26CwrN9Xu7
Davide__Cirillo: RT @Miles_Brundage: "Bias and Generalization in Deep Generative Models: An Empirical Study," Zhao and Ren et al.: https://t.co/26CwrN9Xu7
arxivml: "Bias and Generalization in Deep Generative Models: An Empirical Study", Shengjia Zhao, Hongyu Ren, Arianna Yuan, J… https://t.co/DY0LLMCRiR
BrundageBot: Bias and Generalization in Deep Generative Models: An Empirical Study. Shengjia Zhao, Hongyu Ren, Arianna Yuan, Jiaming Song, Noah Goodman, and Stefano Ermon https://t.co/NrFQi2ooYC
webeneer: RT @Miles_Brundage: "Bias and Generalization in Deep Generative Models: An Empirical Study," Zhao and Ren et al.: https://t.co/26CwrN9Xu7
Miles_Brundage: "Bias and Generalization in Deep Generative Models: An Empirical Study," Zhao and Ren et al.: https://t.co/26CwrN9Xu7
Images
Related