Poster

Improved Autoregressive Modeling with Distribution Smoothing

Chenlin Meng · Jiaming Song · Yang Song · Shengjia Zhao · Stefano Ermon

Keywords: [ generative models ] [ autoregressive models ]

[ Abstract ]
[ Paper ]
Wed 5 May 5 p.m. PDT — 7 p.m. PDT
 
Oral presentation: Oral Session 9
Wed 5 May 7 p.m. PDT — 10:05 p.m. PDT

Abstract:

While autoregressive models excel at image compression, their sample quality is often lacking. Although not realistic, generated images often have high likelihood according to the model, resembling the case of adversarial examples. Inspired by a successful adversarial defense method, we incorporate randomized smoothing into autoregressive generative modeling. We first model a smoothed version of the data distribution, and then reverse the smoothing process to recover the original data distribution. This procedure drastically improves the sample quality of existing autoregressive models on several synthetic and real-world image datasets while obtaining competitive likelihoods on synthetic datasets.

Chat is not available.